Beruflich Dokumente
Kultur Dokumente
..J
C
Cl
'iii
Q)
o
0.7
"0
c
III
CiS 0.05 L..._-L_....I-__.l..---I_....1.--I...-I.....I......l....I.__-I-_...l--_---L_--L----4
0.1
0.4
Natural Period T (s)
Figure-4.4.2 Standard Design Lateral Seismic Coefficient k
hO
for Seismic Coefficient Method
As seen in Table-4.4.1, the ground condition is defined by three types; hard, intermediate
and soft, depending on the N-value obtained by the standard penetration test or the
average elastic shear wave velocity, if available, and the depth of each layer of the soil
above the bedrock. T, the natural period of a vibration unit for design, is obtained by
Eq(4.4.2).
where,
T =2.01.[i
0= Jm(s)u(s)2ds
Jm(s)u(s)ds
(4.4.2)
(4.4.3)
In Eq(4.4.2) and Eq(4.4.3), 0 is the lateral displacement at the centre of mass of the
superstructure when a combination of lateral loads, which consist of 100% of the
superstructure mass and 80% of the substructure mass, is laterally applied. m(s) is the
weight of a mass at the point s, u(s) is the lateral displacement at the point s when all the
m(s) are laterally acting on the column. If k
h
is calculated as less than 0.1 by Eq(4.4.1).
then kh=0.1 is used, instead.
The maximum value for k
h
is 0.3 when Cz=1.0. This implies that the maximum lateral
19
acceleration in case of earthquake, namely the design seismicity in the Seismic Coefficient
Method, is 30% of gravity.
4.4.2.2 Lateral Strength Method
a) General
.Atter the original dimension of a column is decided by the Seismic Coefficient Method,
it is checked by the Lateral Strength Method considering the post-elastic behaviour of the
column. In this step, two types of design earthquakes are assumed, and the column must
be designed to meet the requirement defined for both design earthquakes.
The basic condition of this method is that the lateral strength of the reinforced concrete
bridge column P. defined in 4.4.2.2(d) considering the failure mode of the column, must
meet the requirement of Eq(4.4.4).
where
P a ~ k h e W
W= W
u+
C
p
W
p
(4.4.4)
(4.4.5)
The equivalent weight W is defined by Eq(4.4.5), from the superstructure weight W
u
, the
self weight of the column Wpand the weight coefficient Cpo C
p
is 0.5 if the column fails by
flexure or the column yields by flexure then fails by shear. C
p
is 1.0 if the column fails by
shear. The fracture mode of the column is confirmed in 4.4.2.2(d).
b) Residual Displacement
Bridges are divided into two types depending on their importance. These are
Type A: Not Type B
Type B : Important Bridges (ex. National Highways, Urban Expressways, Over-railway
Bridges, Over-road Bridges, etc.)
If a bridge is regarded as Type B, the residual displacement of the pier OR must be equal to
or less than the allowable residual displacement ORa in order to secure the emergency
traffic and keep it repairable even after a large scale earthquake.
(4.4.6)
ORa is defined as 1% drift, namely 1/100 of the height from the bottom end of the column to
the point where the inertia force of superstructure is acting.
20
The residual displacement OR is obtained by Eq(4.4.7) from the residual correction factor
C
R
, the response ductility factor PR, the ratio of the elastic stiffness of the column to the
secondary stiffness of the column after yielding r and the yield displacement of the
column q,. The design lateral seismic coefficient k
hc
is determined as in 4.4.2.2(c).
where
(4.4.7)
(4.4.8)
c) Design Lateral Seismic Coefficient
As mentioned above, two types of design earthquakes are used and the standard
design lateral seismic coefficient k
hcO
is obtained from each response acceleration spectrum
(Fig-4.4.3).
The Type-1 earthquake represents the earthquakes occurring relatively far out to sea,
caused by the movement of the plate crusts, and the Type-2 Earthquake stands for the
local earthquakes relatively near to the ground surface caused by the movement of faults.
Type-1 spectrum was made from statistically analysed response acceleration spectrums of
394 records obtained in the event of large-scale earthquakes in Japan. Type-2 spectrum
was developed from some seismic waves recorded on different ground conditions during
the Hyogo-ken Nanbu Earthquake in 1995. On the Type-2 spectrum, large peaks of the
acceleration were smoothed, and some adjustments were carried out using a damping
factor for each natural period. Table-4.4.2 shows the standard lateral design acceleration
obtained from those spectras.
2.1
0.2 0.5 1
Natural Period T (s)
!
~
,
~ :
:
i
I i i
!
,
;
! ! !
~
i
i
i
:
i
!
i j
,
;
I
!
I i
2 5
! : : : ;
,' .,f t
'.;,' '
i : ~ :
: Ii
: Ii! :
I I i I I
Soft
i i
, ,
!
, ,
{
f
i
,
t ~
:
i ! i i :
i i
,
i
:
~
;
!
,
,
! !
,
! i
,
,
~ I
i
l }
,
i
i
,
f
,
i ;
!
,
:
,
:
i
:
t i
~
~
:
j
i
I
i
~
:
f
!
-
Hard
:
,
! !
,
- - - - - Intermediate
a) For Type-1 Earthquake
5 2
Hard
- - - - - Intermediate
Soft
2
1
0.5
0.2
O. 1
0.1
0.2 0.5 1
Natural Period T (s)
b) For Type-2 Earthquake
Figure-4.4.3 Standard Design Lateral Seismic Coefficient k
hco
for Lateral Strength Method
22
Table-4.4.2 Standard Design Lateral Seismic Coefficient k
hcO
for Lateral Strength Method
EQtype Ground condition Value of khcocorresponding Natural period T(s)
Hard
1.4<T
k
hcO=0.7
k
hcO=0.876
r
2l3
Intermediate
T<0.18
0.18::o;T::o;1.6 1.6<T
Type-1 k
hcO=1.51
T
1/3
k
hcO=0.85
k
hcO=1.16
r
2l 3
Soft
T<0.29
0.29::o;T::o;2.0
2.0<T
k
hcO=1.51
T
1/3
k
hcO=1.0
k
hcO=1.59
r
2l 3
Hard 7<0.3 O.3::0; T::o;O.7 0.7<T
k
hcO=4.46
T
2I3
k
hcO=2.0
k
hcO=1.24
r
/3
Type-2
Intermediate
7<0.4
0.4::o;T::o;1.2 1.2<T
k
hcO=3.22
7
213
k
hcO=1.75
k
hcO=2.23
r
/3
Soft
7<0.5
0.5::o;T::o;1.5 1.5<T
k
hcO=2.38
T
2I3
k
hcO=1.50
k
hcO=2.57
r
/3
After khcO is chosen from Table-4.4.2 depending of the ground condition and the natural
period of the vibration unit, a design lateral seismic coefficient k
hc
is obtained by Eq(4.4.9)
for each Type-1 and Type-2.
(4.4.9)
Then, the equivalent lateral seismic coefflclent z., is determined for each Type-1 and Type-
2 earthquake by Eq(4.4.10), reducing the design lateral seismic coefficient K
hc
and
considering the ductility of the bridge column J.la based on the Equivalent Energy Theory.
Here, k
he
must not be less than O.4C
z
. The allowable ductility factor of the column J.la is
defined by Eq(4.4.12).
(4.4.10)
d) Lateral Strength of the Column
The lateral strength of the column P
a
in Eq(4.4.4) is obtained as follows.
23
The failure mode of the column is determined from the ultimate lateral strength of the
column Po. the lateral shear strength of the column P, and Pso. Pso is obtained by
Eq(4.4.32) and Eq(4.4.33) assuming the cyclic loading factor c, is equal to 1.0, which
means that Pso is the lateral shear strength of the column before the column is affected by a
cyclic lateral load.
1) If Pu::;; Ps , the column fails by flexure.
2) If P
s
< P
u
::;; Peo . the column yields by flexure then fails by shear.
3) If P sO < P
u
, the column fails by shear.
Then the lateral strength of the column P
a
is determined by Eqs(4.4.11).
1) P = P
u
when the column fails by flexure (Pc<P
u)
2) P = P
u
when the column yields by flexure then fails by shear
3) P
a
= Pso when the column fails by shear
Where Pc is the cracking lateral load.
(4.4.11)
e) Allowable Ductility Factor
The allowable ductility factor of the column lIa in Eq(4.4.10) is determined as
Eqs(4.4.12), using the ultimate displacement of the column ou, yielding displacement of the
column oyand the safety factor a defined in Table-4.4.3.
1) lIa =1 + ( Ou -Oy) I ( a Oy) : when the column fails by flexure;
2) lIa =1 : when the column yields by flexure then fails by shear or when the column
fails by shear (4.4.12)
Table-4.4.3Value of Safety Factor a
Importance of the bridge
a for Type-1 earthquake a for Type-2 earthquake
Type B : Very important 3.0 1.5
Type A : not Type B 2.4 1.2
The cracking lateral strength P yielding lateral strength P
r
yielding displacement or'
ultimate lateral strength Po and ultimate displacement Ou of the column are calculated for
each Type-1 and Type-2 earthquake, based on following criteria.
24
1) A fibre strain in a horizontal section is proportional to the distance from the neutral axis of
the section to the fibre.
2) The relationship between lateral strength and lateral displacement of the column is bi-
linear as shown in Fig-4.4.4.
Lateral
Load
p. .
yo
Figure-4.4.4 Bi-Iinear Model relationship between lateral strength and lateral displacement
of a reinforced concrete bridge column
3) The stress-strain curve of concrete is determined considering the confining effect of
transverse steel.
4) The stress-strain curve of a reinforcing bar is bi-Iinear, as shown in Fig-4.4.5.
Steel Stress 0.
Steel Strains,
where;
o.y: Yield Stress of Steel (kgfI em?')
0. : Steel Stress (kgf'/crn")
E. : Modulousof Elasticity (kgf/em?)
e , : Steel Strain
Figure-4.4.5 Strain-stress relationship of reinforcing bars
25
5) The yield point is defined as the end point of the elastic part of the bi-Iinear model shown
in Fig-4.4.4.
6) The ultimate limit state of the column is defined as the point where the strain of the fibre
at the extreme re-bars in compression reaches to the ultimate strain of the concrete. The
ultimate displacement of the column is obtained from Eq{4.4.13), considering a plastic
hinge forms at the base of the column, from curvature at the bottom section of the
column u and y, plastic hinge zone length L
p
and height of the loading point from the
bottom of the column h. u and y correspond to the curvatures at the ultimate limit state
and the yield point of the column, respectively. 0 is the lateral width of the column.
where
Ou::: Oy + ( u - y) L
p
( h - L
p
I 2 )
Lp=0.2h-0.1D (0.1 D ~ L p ~ 0 . 5 D )
(4.4.13)
(4.4.14)
The lateral strength and lateral displacement of the column at the cracking point, first
yield point, yield point and ultimate limit state are calculated as follows.
1) The column' is equally divided into m horizontal layers (the specification recommends
about 50), then bending moment and curvature for cracking and first yielding are
calculated. The bending moment and curvature for cracking point, Me and (Pc,
respectively, can be calculated by (4.4.15) and (4.4.17) respectively.
where
Me::: ~ ( 10 O"bt + N
i
I Ai)
- 0 5 213
O"bt - O"ek
(4.4.15)
(4.4.16)
(4.4.17)
The elastic modulus and the area of the main bars are considered when calculating ~
(the section modulus) and I; (the second moment of inertia at a section I). N,and Aiare
the axial load and the sectional area at a section i, respectively. The tensile strength of
the concrete O"bt is calculated by Eq{4.4.16) from the design strength of the concrete O"ek.
Each layer is again divided into n fibres (suggested about 50), then the location of the
neutral axis of each layer is calculated by trial (goal seeking) to achievethe equilibrium of
Eq{4.4.18), assuming that the strain of each fibre is proportional to the distance from the
neutral axis.
26
(4.4.18)
Where, O'ej and O'sj are stress of the concrete and reinforcement in a fibre j, respectively.
M
ej
, M
sj
are sectional area of the concrete and reinforcement in a fibre j, respectively.
After the location of the neutral axis is determined, the bending moment M
i
and the
curvature i at the section i are obtained by Eq(4.4.19) and Eq(4.4.20) respectively.
(4.4.19)
(4.4.20)
Where, Xj is the distance from the neutral axis to a fibre t, SeOis the strain at the edge of
concrete in compression and Xo is the distance from the edge of the concrete in
compression to the neutral axis of a layer i.
The bending moment and curvature, when the strain of the longitudinal bar which is in
tension and located farthest from the neutral axis reaches to the yielding strain Ssy, are
calculated and defined as initial yielding moment M
yo
and initial yielding curvature yo,
respectively. The bending moment and curvature obtained when the strain at the fibre
with the longitudinal bar, which is in compression and located farthest from the neutral
axis, reaches the ultimate compressive strain of concrete Seu are defined as ultimate
bending moment M
u
and ultimate curvature u, respectively (see Fig-4.4.6).
2) The initial yielding displacement ~ o can be calculated by Eq(4.4.21) using the curvature
distribution along the vertical axis of the column when the yielding lateral strength Pya
acts at the level of the superstructure mass.
m
== L: (IYI +1-1YI-1 )L1YI /2
1=1
(4.4.21)
3) The reference yielding curvature r and the reference yielding displacement ~ are
obtained by Eq(4.4.22) and Eq(4.4.23), respectively.
27
y= ( Mul Myo) ,pya
Oy= ( Mul Myo) Oya
(4.4.22)
(4.4.23)
4) The ultimate lateral strength P
u
is calculated by Eq(4.4.24).
(4.4.24)
5) Fig-4.4.6 shows the distribution of curvature over the column and Fig-4.4.7 shows the
relationship between the plastic hinge zone length Lp and aspect ratio hiD.
~ I I ~
(Ultimate State for Type-2)
I. ,pur .1
(Ultimate State
for Type-1)
\,pc
... ':-..
\\
.,
~ I
~
(Initial Yielding)
(Reference Yielding)
Figure-4.4.6 Curvature Distribution over Column
8 7 1 234 5 6
Aspect Ratio n/t:
Figure-4.4.7 Plastic hinge zone length
:5 0.8 r--,-.....,---,.-.....,..---r---.,..--..----.
Cl
55 0.7
...J
(l) 0.6
c
~ __ 0.5
(l)Q
ClX 0.4
c.......,
:c . . . . . : j ~ 0 . 3
o
1;) 0.2
CO
0.. 0.11--........(
ol..---L-_.l.-----L.._...L---l._-L----l._-'
o
28
f) Strength of Confined Concrete
The stress-strain relationship of confined concrete shown in Fig-4.4.8 is defined in
Eq(4.4.25)
or
Where
(Tee = (Tek + 3.8 a Ps (Tsy
033
Ps(Tsy
Gee = 0.002 +O. 13--
(TCk
2
Ec!eS = 1 1 . 2 ~
Ps(Tsy
( Gee < Ge :::; Geu) (4.4.25)
(4.4.26)
(4.4.27)
(4.4.28)
(4.4.29)
Geu =Gee (Type-1 Earthquake)
or Geu =Gee + 0.2 (Tee! E
des
(Type-2 Earthquake)
4A
Ps = __h :::; 0.018
sd
(4.4.30)
(4.4.31)
The stress and the strain of the concrete are (Te and Ge; (Tee is the maximum strength of the
confined concrete; Gee is the strain in the concrete corresponding to (Tee; Geu is the ultimate
strain of the confined concrete; E
des
is the decreasing angle of the stiffness of the concrete.
The volumetric ratio of transverse reinforcement is Ps; the vertical spacing is s and the
effective length d is the shortest length of a confined concrete section surrounded by the
transverse reinforcement, as seen in Fig-4.4.9. Sectional adjustment coefficients a and 13
are equal to 1.0 for a circular section column, a. is equal to 0.2 and p is equal to 0.4 for
square, rectangular, hollow circular, hollow square and hollow rectangular section columns.
The coefficient n is defined by Eq(4.4.26). As seen in Eq(4.4.30), ps should not exceed
0.018 because too much confinement generally results in a shorter plastic hinge zone
length and early fracture of main bars under cyclic loading.
29
Stress
i
i
..._._--....- ..-!----_......_.
I
I
{
I (e )1I-1}
a=Ee 1--_
c
C cc n t: i
"cc i
:
Strain
Figure-4.4.8 Stress-Strain Relationship of Confined Concrete
d
(a) Circular Column
...
"1:S
-e
+
-e
~
~
s: Direction: d=Largest among
dl-d
a
YDirection: d =Largest among
dj-d
s
(b) Rectangular Column
XDirection: d =Largest among
d.-d
s
Y Direction: d =Largest among
ds-d
lJ
(c) Rectangular Hollow Column
Figure-4.4.9 Effective Length of Transverse Reinforcement
30
g) Shear Strength of the Column
The shear strength of the column P is calculated by Eq(4.4.32), as a sum of the
concrete shear strength Se and the transverse reinforcement contribution Ss. The concrete
shear strength Se is obtained from the average concrete shear strength ZOe, seen in Table-
4.4.4, and the transverse reinforcement shear strength Ss is obtained by the yielding stress
of transverse reinforcement O"sy'
P
s
=s, + Ss
s, =10 C
e
c, Cpt ZOe b d
S =AwO"syd(sine +cose)
s 10x1.15s
(4.4.32)
(4.4.33)
(4.4.34)
If the column height h
p
is less than d/1.15, h
p
is used instead of d/1.15 in Eq(4.4.34). The
cyclic loading effect adjustment factor C
e
is equal to 0.6 for Type-1 earthquake and 0.8 for
Type-2 earthquake. This difference describes the number of large waves during an
earthquake. Usually, the number of critical waves is relatively large in the case of a large
earthquake, which Type-1 earthquake represents, occurring far out to sea due to the
movement of crustal plates. On the other hand, the number of critical waves is relatively
small in the wave record of Hyogo-ken Nanbu Earthquake on which Type-2 earthquake is
based. The size effect adjustment factor C
e
is verified by the effective depth of the column
d as seen in Table-4.4.5. Cpt is an adjustment factor concerning the volume ratio of main
bars in tension, Pt, seen in Table-4.4.6.
Table-4.4.4 Average Concrete Shear Strength ZOe (kgf/cm
2
)
Design Strength of Concrete O"ek
210 240 270 300 400
Average Concrete Shear Strength ZOe
3.3 3.5 3.6 3.7 4.1
Table-4.4.5 Size Effect Adjusting Factor C
e
Concerning d, Effective Depth of the Column
Effective Depth d(m) ::;;1 3 5 10::;;
C
e
1.0 0.7 0.6 0.5
Table-4.4.6 Adjusting Factor Cpt Concerning Pt
Volume Ratio of Main Bars in Tension Pt (%) 0.2 0.3 0.5 1.0::;;
c;
0.9 1.0 1.2 1.5
31
h) Reinforcement Details
In order to secure enough ductility of a reinforced concrete bridge column, the following
reinforcement details are required to be met.
1) No joint in the main reinforcement shall be located within 4 times the plastic hinge zone
length L
p
, determined by Eq(4.4.14), measured from the base of the column.
2) Hoops should be deformed re-bars of diameter equal to or more than 13mm, the
centre-to-centre vertical spacing of transverse reinforcement should be equal to or less
than 150mm, and the centre-to-centre vertical spacing can be expanded gradually up to
300mm higher up the column, if necessary (e.g. a tall-pier, more than 30m.)
3) A hoop must be anchored in the concrete core of the column by hooks and each must
surround a main bar. gO-degree hooks should not be used at a comer of a rectangular
or square column. Location of the hooks should be staggered along main bars. The
inside-bending radius of a hook must be 2.5 times the diameter of the bar.
Development of a hook must be at least 8 times the diameter or 12cm for a semi-
circular hook, 10 times the diameter for 135-degree hook and 12 times the diameter for
a so-decree hook.
4) Unless the two ends of hoops are hooked at a comer bar of a rectangular or square
column, they must be spliced by more than 40 times the diameter of the bar and each
end must have a hook defined in (3).
5) For rectangular, square or wall type sections, hoops should be arranged surrounding
main bars and cross bars should be also used. Cross bars are designed as follows.
a. Diameter and strength of cross bars should be the same with that for hoops.
b. Cross bars should be set parallel to the edges of the column section.
c. Horizontal spacing of cross bars, namely the effective length of transverse
reinforcement d, should be equal to or less than 1m.
d. Every hoop must have cross bars.
e. Cross bars must hook the hoop in the same section. If the main bars are double
layered, cross bars can hook only the outer hoop.
f. Hooks of cross bars hooking hoops should be 135-degree hooks or semi-circle
hooks, defined in (3). If gO-degree hooks need to be used due to some difficulty in
construction, only one side of a cross bar can have a gO-degree hook. In this case,
1.5 times the greater value for effective length of transverse reinforcement d must be
used in Eq(4.4.31), in the region from the base of the column to the height of 4 times
the plastic hinge zone length L
p
Location of gO-degree hooks must be staggered.
g. Any cross bar should not be spliced in the column. If necessary, a joint that can
generate the same tensile strength as the bar must be used.
32
6) Hoops in a circular column must surround all the main bars. Cross bars should be
used also in a circular column mainly for better shear resistance, not for confinement.
The length of a splice joint in a hoop must be more than 40 times the diameter of the
bar. Both hoops and cross bars must have a hook defined in (3) on each end.
7) A hollow section column must have hoops and cross bars to confine its concrete
sections. Detail of these cross bars must meet the requirement for rectangular or
square sections defined in (5).
~ 0 c
v v u v
c
::> (
(
o
c
D
(
P
\
P
c
p
n ~
h t"\ '" '" n
a) Square Section b) Wall Type
c) Circular Section
d) Rectangular Hollow Section
Figure-4.4.10 Standard Layout of Transverse Reinforcement
33
Figure-4.4.11 Examples of Joints for Cross Bars
Main bars should not be curtailed at any part of a column, unless the column is
extremely tall (e.g. over 30m). If necessary, the following requirement must be met. The
main bars must not be curtailed in the region from the base of the column to a height of 4
times of plastic hinge zone length L
p
The number of the main bars curtailed at one section
must not be more than 1/3 of all the main bars. Centre-to-centre spacing of transverse
reinforcement must be equal to or less than 15 cm around the curtailment point of main
bars, from 1.50 below the curtailment point to 1.50 above the curtailment point. The
centre-to-centre spacing of transverse reinforcement must not be changed suddenly.
34
Chapter V
Test Progran1l
5.1 General
Originally, five identical test specimens were planned to be tested with five different
loading patterns. However, the strength of the concrete cast in one column was found to
be less than half of the design strength due to an error in manufacturing the concrete, as
seen in 6.2.2. Therefore, the original plan was revised to test four identical test
specimens with four different loading patterns and one weak-concrete column with the
standard loading pattern.
5.2 Testing Procedure
5.2.1 Test Set-up
The test specimens were supported between two universal joints at the top of the
column and the bottom of the base block, and axially loaded by a DARTEC universal
testing machine. A hydraulic jack in the floor applied an axial load to the specimen pushing
up the specimen with the loading rig and the counter weight. The hydraulic jack was
controlled by a computer and automatically maintained a uniform axial load on the
specimen. An L-shaped loading rig was fixed to the base-block using nine high-strength
bolts horizontally penetrating the base-block in each direction, and a counter-weight basket
was set on the opposite side of the rig using the same bolts. A 900mm
x900mmx750mm
concrete block (1.5tf) was set in the basket on the opposite side of the loading rig as a
counter-weight to the rig, horizontal jack and other parts. A horizontal hydraulic jack was
connected to the top of the loading rig and to the loading point of the column which was
50cm below the top of the column. One set of loading rig with counter-weight basket was
used for the specimens-1, 2 and 5 in the east-west direction, and two sets were used for
specimens-3 and 4 in east-west and north-south directions in order to apply bi-directional
horizontal load. When the horizontal jack applied a displacement to the column, flexure
was applied at the base of the column resulting in a rotation of the loading rig, base-block
35
1
2475
I
_L
' ..... ' AA
I
I\."'"
- - - - - -- ,
I
t
I Hydraulic Jack
I
c;L _ Column
0
I
-
Loading Rig_
I.J"I
N
I
......
~
N
I
I
I
Base Block
I
I
Spacer Block
Counterweight
\
i
\
/
Elbow
,
1
Joint
I
I
~ ' /
V
1/
,5=
-,
I
I
Ir-
-
: I
v '/
2440
QM <t :tc;o
2370
(J,)'
en
CGUnter Weight Basket
EastlSouthl
Figure-5.2.1 Test Set-up [unitmm]
Test Rig
WestlNorthl
--
0.>
:::r
:::l
CD
0-
o
o
0
0
:::l
C
o
:::l
CD
--
'"tJ CD
--
.....
sa. ::E
--
CD
:::r
cO'
CD
:::r
-- --
CD
e-
en
0.>
-- en
en
"
CD
~
--
I
C
::E -o
:::r
c:r
:::r
::E
CD
CD
o
0
:::l
:::l
CD
~
CD
0-
0
CD
0.>
o
:::r
sa.
:::r
CD
.....
S'
~ ,
:::l
CD
" cO'
c
@
,
01
N
......
en
zr
~
en
5.2.2 Loading Patterns
The main objective of this experimental study was to estimate the difference in damage
on a reinforced concrete bridge column caused by different loading patterns. Four different
loading patterns were applied to identical specimens, Specimen-1 to Specimen-4 as follows
in 5.2.2.1 to 5.2.2.5.
5.2.2.1 Specimen-1
Specimen-1 was loaded by the standard loading pattern used at the University of
Canterbury seen in Fig-5.2.2, described in 4.2 (Park [2]), as a bench mark test. The test
specimen was considered to had failed when the lateral strength of the column was smaller
than 80% of the maximum lateral load measured during the first cycle to 2OY. This
definition of the ultimate state of a specimen is the same as that Zahn et aJ. [18] used, and
applied to also Specimens-3, 4 and 5. For all the rest of this research report, "ultimate
state" means that defined by Zahn et aJ. [18].
5.2.2.2 Specimen-2
Fig-5.2.3 shows the pattern used for Specimen-2 which was completely opposite to
that used for Specimen-1, except for the first cycle to three-quarters of the ideal strength
used to determine the reference yield displacement oy. The maximum displacement of the
second and third cycle was 12oy, which was derived from the test result of Specimen-1,
then the applied maximum displacement for a loading cycle was decreased by 20y in every
two cycles. The test was terminated after two cycles of 2OY.
37
10 :......... . .. ... ,'lLJ;II, -11- :-1
2 " ) 4
5
,2 6
, '
-5 ... .. .. -.. ---_. - ...,-_._-- - .. -----
, '
, , ,
::: t11llfl1 .. u- e- -.- -.-
"'1W
3-
.....
c
Q)
E
Q)
c
I'll
0.
.!a S
o
Step
Figure-5.2.2 Loading pattern for Specimen-1
15
1
G ' 4 ',8 :2
...----- ..,.-----
--------,. ------u- -:.- ---.. -D-,- --- . -lI.; -. -- -.110;. -- -.--- --.--..i. --. -. -.). - J ; -. --.i. l.- -.. -
, , , , " 't
, , 1 , " "
I , , , " "
: i I I ,I : I
10
-10
, , , , ,
-- ,--II---:-ft---;--.,-+1ID +++' ';1
"; w 5 +.-----.-:-1 --., --- .. - - -.-'.- ---:- -----:---r------ ---r-----
3- :: :
......, J I
c: :: :
Q) :: :
E 0
s
I'll
0.
is S -5
Step
Figure-5.2.3 Loading pattern for Specimen-2
38
5.2.2.3 Specimen-3
A bi-directional loading was used for Specimen-3. The standard loading cycle, same
as that for Specimen-1, was applied for each north-south and east-sest direction in one
direction at a time. Specimen-3 was loaded up to three-quarters of the theoretical ideal
strength in each of the four directions; east, west, south and north in tum. Four reference
yield displacements for all directions were calculated from the measured displacements and
the average value of them was determined as the reference yield displacement q, of
Specimen-3. Then two cycles of 2q, were applied in the east-west direction, and after the
column was returned to the initial position (zero-displacement), the same two cycles were
applied in the north-south direction, l.e. a total of four cycles of 2q, were applied. Two
cycles of 4q, in the east-west and two cycles of 4q, in the north-south direction followed
subsequently, and the displacement amplitude for a loading cycle was increased by 2q, in
every four cycles.
5.2.2.4 Specimen-4
A bi-directional loading was used. The procedure to determine the yield displacement
q, was the same as that for Specimen-3, but for rest of the loading cycles, the previous
direction of loading was held at the maximum reached displacement while loading in the
new direction was applied. The loading amplitude was increased by 2q" in every one
complete cycle seen in Fig-5.2.5(c).
5.2.2.5 Specimen-5
The same loading procedure as forSpecimen-1 was used also for Specimen-5, in
order to compare the result to that from Specimen-1.
39
, , .
--_.- .. ----.. -----
, , .
,
,
, , ,
6an
.- - -----;.-----:-----
_____ '- .J _
, , ,
J .'- .J __
, , ,
, , ,
, , ,
, , ,
, . ,
_._w'. . _. . -----'. . .'_.....-
- - - .. - - - - _. - _. - ... - - _ - - ---.- - - - _- - - .j- - ---. - - ---... - -_ .. - - - - - - - - .. - - -- ... --
, , , , , 1 , , ,
, , , I , ,
, ",' I
I "",
, " "., '"
- - _..,_ - - - -, - - - -..,- - - - - .. - - - - '0- _. - - - - - - 't - - - - - - _. -,- - - - - - - - - - .. - - - - ... -'r - , -.
, """"'" I
, I' . , , , , ,
, ,. , , , I , ,
, " , , , , , ,
" , I , , , ,
, ,
__w..: ..: .. - . ..:- _
, ,
, '
----:----- ---- ---- -----,----- ---- ---- -----,----- ----
10
8
6
4
2
o -1'-'0,---i-- 1--\-;-'i"-'04-1H -1- ....................-t t--HI"""""......+ IH-+..."""""F--.ot-
-2
-4
-6
-8
-1 0 ..1..-...:.---'---'---'-'---'---'---'---'---'----'-"---'---'---'=-----------'---'
'";w
3-
......
c
Q)
E
Q)
U
ell
0..
.!Q ::;
o
Step
a) Loadgin Cycle for East-West Direction
--
, ,
, . .
,-- :----,-----:---IIi-U:-----j
, , ,
. . ,
, , ,
, , ,
, , , ,
, , , ,
, , , , , , " ,," ,
---_:_ ---- ;" .. -_! -- ;_ --_. ----_:_. ---:__ --_; - ; -_: --:_ -_. _:' _.. ";__ 0- ow:
" ",'. , " ,
" """" ,
, " """'" , __ ... - - -...,_. - _ - - - -,- _. - T - _ ..... T - - - -'r - - - - .... _. -,. - - - -... - - _.. _ ...._ _. -'1
, , , " , '" t
, '" , '" I
, ." , '" ,
, , '" ,,',' I
----:-----r----r--
t
-----r ----r----r ----1 --y----r-----j _.y---- ow:
I ' j , , , " "
_ .,. _ - _ - - - - .... - _. - -.. - - .j- _ T - _. - - ..... - -.. - -.., - - -,. - - - - - _ - _ ",- _ - - - __ --
" "" " I
", , ,
. , ,
, , ,
+IH-,.....""""""'........IH
" ,
, " , "
---: -. ---} ----T----! ----T' --. -----r ----1- ---':- ----:-----r-----:- ----t--. -T-. --(- ----;- --
10
8
6
4
2
-2
-4
-6
-8
-10 _ ___'___'-'-___'---'___'___'_ ___'..:.._...:.___'___'____'___''---J
'1Cf)
3-
......
c
Q)
E
Q)
U
ell
0..
is z
Step
b) Loading Cycle for North-South Direction
12,16
U11,15
3,7 2,6
:::::J
<===:J
...
...
c:::::::>
c::=::>
4,8
n
U 1,5
c) Loading Orbit at the Top of the Specimen
9,13
10,14
Figure-5.2.4 Loading Pattern for Specimen-3
40
- _. -.,.> - _. - - - - - --
, ,
--------. - _-_. -----,.-.-- -----r----
, ,
, ,
, ,
, ,
--_.:._._. -----;--_. _...
... -:.. _-- ..... : .. _-
10
8
6
4
3-
2
-<oJ
l::
Q)
E
0
Q)
o
ro
-2
c..
iSs
-4
-6
-8
-10
Step
a) Loading Pattern for East-West Direction
-. ---_.,----'
-----;-----
, , ,
,-----,.0 .. _..
, , ,
, . .
- ,-----1"-_...,- - -
. .
, , .
. .
, .
, ,
- -----1'-----, - -
-- --.. -1'--- -- --
-----,----_., --
- - - --:- -- _. -
- - - 't' - - - -: - - - _. '- - - - - -:- _ - - - - - - - +----- ----";"--- -";
" I , ,
" , , ,
" " .",
. ---:- - - -- --- _. ---- -;- _. --":- ----: -----;- -----;- ---- :-----;". - - --;
, , " "
, , " ,
, , , , " "
....: ; : : , : : : : : : : :: jI..
, , , , " I
" ," ,
" ,,' ,
10 -r------:----:--:--;--:---:--;-;-:--:--;--:---:--;-;---;---,
8
6
4
2
o
-2
-4
-6
-8
-10 "'-- --"" ....
"J(j)
3-
-<oJ
l::
Q)
E
Q)
o
ro
c..
6 z
Step
b) Loading Pattern for North-South Direction
A
....
...
<;::::==J
All.
<;::::==J
All.
6
11
U
7 5lJ
n
12
9
1tl
8
,I. c=:::> c=:::>
..
l 0:
c:::=:> c:=::>
16
1
lJ15
14
4tl
n13
3
2n
<;::::==J
<===:J
c) Loading Orbit at the Top of the
Specimen
Figure-5.2.5 Loading Pattern for Specimen-4
41
5.3 Design of Specimens
5.3.1 Design Process of the Specimen
The test specimens are models of a reinforced concrete bridge column with a square
section, designed by the Japanese specification. The prototype column is a design
example in "Seismic Design Data for Highway Bridges"[19], which was published
immediately after the latest version of "Specification for Highway Bridges"[1] in order to
explain the usage of the specification by some design examples. The design of a typical
Japanese highway bridge is used for the design example in the book. The original column
cross section of the design example is 5mx2.2m, but this was modified to a 3mx3m square
section then scaled down to the test specimen with 550mmx550mm section. A square
section was chosen for the specimen in order to reduce the number of the parameters in
the experiment and its analysis, and to make the application of the results more general.
The design conditions for the prototype bridge is explained in 5.3.2 and is derived from the
design example. The design procedure for the prototype column with a rectangular section
follows in 5.3.3 using the design process of the Japanese specification explained in 4.4.
The seismic design of the column is carried out in the longitudinal direction and the
transverse direction, separately. The prototype rectangular section is modified to a square
section in 5.3.4 and scaled to the test specimen size in 5.5.5.
5.3.2 Design Conditions
The prototype bridge, seen in Fig-5.3.1, has five continuous spans with I section beam
supporting a reinforced concrete slab. Rubber bearings are placed between the
superstructure and each pier and abutment. Each span of the superstructure is 12.0m
wide and 40m long. The superstructure is laterally fixed in the transverse direction at both
abutments but elastically supported in the longitudinal direction. The superstructure is
elastically supported in any lateral direction on every pier.
The column of the pier P1 is chosen as the prototype column of the specimen. All the
piers are T-shaped and have cast-in-place reinforced concrete piles, as seen in Fig-5.3.2.
The design strengths of the materials are seen in Table-5.3.1.
42
200000
5X40000=200000
r -DesignedPierl
:Elastic Bearing
(a) Elevation
I
I 0 CO
I 0
0
I
I
0
I
0 0
N N N
N N N
-
r"'''"'
- -
,,""w"",
-
I
I I
I I Inpermiable
---
____L_ !-_L _____
-------- -------- -------
__
3000
12000
8000
-n
1000 4X2500=10000 1000
o
o
N
N
(b) Superstructure Section
Figure-5.3.1 Elavation and Sectionof the Prototype Bridge [unitmm]
12000
o
o
N
N
3150 3150
2200
Sup" "NOU"O Mass rr--;
3500
6000
50'00
Super structure -'-
Mass I
1
3500
6000
o
o
o
'"
N
o
:;;:1----+-----""-,
o
o 0
o
'" U1
- r-.
Cast in Place
Piles 91200
L=15.00m 0=9
(a) Front Elevation
(b) Side Elevation
12000
3500 5000 3500
I
o
-$
/l--
-$
U1
T
-
'" 0
0
'"
0
N
N
0
U1
Jr
0
0
'"
-$ -$
U1
<'l C
\!,.
;;:' i
-
I
I
120C 3050 I 3050 120h
8500
o
o
U1
co
(c) Plan
Figure-5.3.2 Reinforced Concrete Bridge Pier Designed (P1) [unitmm]
44
Table-5.3.1 Design Strengths of Materials
Member Concrete Reinforcing bars
Pier f'c=210kg/cm" (20.6MPa)
Footing f'c=21 Okg/cm"(20.6MPa) SD295 (fy=295MPa)
Pile f'c=240kg/cm" (23.5MPa)
Table-5.3.2 shows the superstructure loads on the substructure parts, used in the
analytical frame model of the prototype bridge, seen in Figures-5.3.3 and 5.3.4.
Table-5.3.2 Vertical Load on Substructures
Substructure Dead loads RD Live loads R
L
A1,A2 abutments 250 tf ( 2450 kN ) 140 tf ( 1372 kN )
P1,P4 piers 710 tf ( 6958 kN ) 260 tf ( 2548 kN )
P2,P3 piers 610 tf ( 5978 kN ) 250 tf ( 2450 kN )
The prototype bridge is in a region with a high probability of a large scale earthquake, and
the importance of this bridge is ranked as high, namely Type-B defined in 4.4.2.2{b). The
ground conditions are assumed to be "intermediate soil" and liquefaction is not considered.
5.3.3 Design of the Prototype Column
Based on the design condition shown in 5.3.2, the prototype reinforced concrete bridge
column is designed as a wide rectangular column, in "Seismic Design Data for Highway
Bridges"[1] as follows.
5.3.3.1 Seismic Coefficient Method
Two-dimensional frame models of the bridge are used to obtain the natural period of
the vibration unit for the prototype pier P1, which in this case is the whole of the bridge,
based on the seismic coefficient and the superstructure weight on the designed pier. Static
analysis programs are generally used to process these models. Two frame models of the
prototype bridge; one for the bridge-axis direction and the direction perpendicular to the
bridge-axis, are shown in Fig-5.3.3 and Fig-5.3A.
45
20000 20000 20000 20000 20000 20000 20000 0000 20000 20000
2 6 7 8 9 10 11
"" 101 301 401 2001
-!l
102
1
302 402
-0)
C")
2002
1031
I
303 403
0
I I
If)
....
I I
2003
C")
1041 I
204 304 404
2004
I I
2005
I I
I I
1O..J
L..,
205 305 405
r,06 206 306 406
I I
407
.J:>.
tl.0.l 207 307
(j)
<AD ) @ <)
Z
.: Node with Mass
0: Node where column section changes
~ x
Note: 1 and 1001, 3 and 101, 5 and 201,7
and 301 have the same locations.
Superstructure is fixed in Z direction,
elastically supported in X direction, and
free along Y axis, at elaslicbearings.
Figure-5.3.3 Two-dimen sional Frame Model (Longitudinal Direction)
20000 20000 0000 20000 20000 0000
2 3 4 5 6 7 8 9 10 11
0
100 200 300 400 0
to
2001
201 301 401
202 302 402
2002
I
203 303 403
to I
l"-
I
2003
C'")
I 204 304 404 2004
I
2005
0
I UJ
r-
I <')
~
L-,
205 305 405
-..I 0
206 306 406
I
o ~
207 307 407
z
tL-x
: Node with mass
0: Node where the columnsection changes
Note: 100 and 101, 200 and 201,300 and 301,
400 and 401 have the same locations.
Superstructure is fixed in X direction,
elastically supported in Y direction, and free
along Z axis at elastic bearings.
Figure-5.3.4 Two-dimensional Frame Model (Transverse direction)
The whole section of any member is assumed to be effective for bending moment, at this
stage. Horizontal displacement Vi of each node i caused by the lateral load equal to the
self-weight is calculated using these frame models. From these displacements Vi and the
weights of each node, Wi, the natural period of a vibration unit for design, T, is obtained
using Eq(5.3.1) derived from Eq(4.4.2) and Eq(4.4.3).
T = 2.01 / - = - 7 - ' - ~ (5.3.1 )
The natural periods of the prototype pier P1 for each direction are obtained as follows,
assuming the section seen in Fig-5.3.5.
T=1.15(s)
T= 1.02 (s)
for longitudinal direction
for transverse direction
Therefore, the design lateral seismic coefficient k
h
for the seismic coefficient method is
determined from Table-4.4.1 and Eq(4.4.1), for both directions.
K
hO
=0.25
K
h
= C
z
khO = 1.0 x 0.25 =0.25
Also, the superstructure weight supported by the designed pier P1 is obtained from the
analysis.
W
u
= 643 (tf)
':\,W
u
=399 (tf)
for longitudinal direction
for transverse direction
The specification requires the minimum weight to be carried equal to 3/4 of the total dead
weight acting on the designed pier, for both directions. From Table-5.3.2, the minimum
weight is obtained as 0.75x610=474 (tf). Table-5.3.3 shows the calculated natural
periods, design lateral seismic coefficients for seismic coefficient method and
superstructure weight for P1.
48
Table-5.3.3 Natural Periods, Design Lateral Seismic Coefficients for Seismic
Coefficient Method and Superstructure Weight Supported by P1
Longitudinal Transverse
Natural Period T(s) 1.15 1.02
Design Lateral Seismic Coefficient kh 0.25 0.25
Superstructure Weight W
u
(tf) 643 474
Fig-5.3.5 shows the assumed section at the base of column P1, and Table-5.3.4 shows the
check result of the Seismic Coefficient Method.
5000
1201 0 36X125=4500 130 PO
Vertical
Vertical
Main-Bars 0 2 9
Main-Bars 0 2 9 Hoops 0 13 etc 15 0
~
\\\\ .....
~ N
U') - ~
~
0
~
0
0
""-
co
LD
::::
0)
t-
o
~
::::
0
II
N
LD
N
N
~
X
0
<r
~
co
0)
LD
~ -
. .
....
,-, N
N
I I
~
~
Cross-Bars 0 13 e t e 15 0
Cross-Bars
D13et elSO
12b 880 3X1000=3000
I 880 12n
Figure-5.3.5 Bottom Section of P1 Designed by Seismic Coefficient Method [unitmm]
49
Table-5.3A Check Result for Bending Moment and Shear Force at the Base of Pier Pi
Longitudinal Transverse
Direction Direction
Sectional Force Flexure tt-rn 2114.7 1992.2
Axial load tf 1056.2 1056.2
Shear tf 247.6 205.6
Dimension b cm 500.0 220.0
h cm 220.0 500.0
do cm 12.0 12.0
d
,
cm 208.0 488.0
Design Strength of
O"ek
kgf/cm" 210
Concrete
Yield Strength of
O"sy
kgf/cm" 3000
Reinforcement
Tension Steel Ratio
Pt
% 0.334 0.323
Steel Quantity As cm" 250.536(D29
a)*39)
109.208(029*17)
Moment Stress
O"e
kgf/cm' 83 37
o"s
kgf/cm" 2517 495
Allowable Stress 0)
O"ea
kgf/cm" 105 105
O"sa
kgf/cm" 2700 2700
Average Shear Stress rm
kgf/cm' 2.4 1.9
Allowable Shear Stress C
e -
0.838 0.606
Cpt - 1.034 1.023
rat
kgf/cm" 3.3 3.3
rae e)
kgf/cm' 2.9 2.0
Judgement
- rm::; rae OK rm::; rae OK
a) Deformed bar with diameter of 29mm
b) Allowable stress is obtained from a table in the specification
c) rae = C
e
Cpt rat
b
o
-01
.. II .. ., ..
[----t
As'
-u
..c
As
It .. .. .. ..
0
50
In Table-5.3.4, it is seen that each bending moment stress is less than the allowable stress
for both the concrete and reinforcement, in both directions. Because the shear stress is
less than the shear capacity of the concrete, transverse reinforcement is designed only to
meet the minimum requirement.
Finally, the minimum requirement of the reinforcement is checked. But it is not
described here, because a flexure column like the prototype column, designed to have
some ductility, will have more reinforcement than the minimum requirement.
5.3.3.2 Lateral Strength Method
a) Natural period, design lateral seismic coefficient and superstructure weight
Because the section for P1 assumed in the Seismic Coefficient Method, seen in Fig-
5.3.5, does not satisfy the requirement of the residual displacement by Eq(4.4.7) in the
Lateral Strength Method, new column reinforcement is designed as shown in Fig-5.3.6.
The diameter of the main-bars is changed from 29mm to 32mm and the main-bars for the
longitudinal direction of the bridge are changed from one-layered to two-layered. Moreover,
the diameter of the transverse reinforcement is changed from 13mm to 16mm. Here,
however, only the check procedure for the section in Fig-5.3.6, is explained in this chapter.
The two-dimensional frame models shown in Figures-5.3.3 and 5.3.4 are used again to
determine new natural periods, design lateral seismic coefficients and superstructure
weights, using the new bending stiffness of the piers. The bending stiffness K
y
of the
column is determined as a ratio the yielding lateral load of the column P
y
to the yielding
lateral displacement of the column 4 (K
y
=P
y
I 4). Because the bending stiffness
changes as the column dimensions or the amount of column reinforcement changes, the
stiffness of the finally determined section is used here.
51
5000
1? n1 0 36X125=4500 130 1
bO
Vertical Vertical
Main-bars 0 3 2 Main-bars 0 32
Hoops 0 16 etc 150
0
~
0
o,
_ I ~
-
I ~
I ~ . r .
~
-
c
-
0
f::::
0
0 eo
U1
r-, eo
0
-
t::
0
II
N
U1
N
N
-
X 0
N
eo
-
ec
0
n
- - -
~
o_
r
-
J
I I
C\:
-
Cross-Bars 0 16 etc 150 Cross-Bars 0 16 etc ,0
880 I 3Xl000=3000 880 121
Figure-5.3.6 Section at Base of P1 Designed by Lateral Strength Method [unit:mm]
The new natural periods of the pier, corresponding to the design lateral seismic coefficients
for the Lateral Strength Method k
hc
, from Eq(4.4.9) and superstructure weights are obtained
for each direction as shown in Table-5.3.5.
Table-5.3.5 Natural Periods, Design Lateral Seismic Coefficients for Lateral Strength
Method and Superstructure Weight Supported by P1
Longitudinal Transverse
Natural Period T(s) 1.18 1.04
Design Lateral Seismic Type-1 EO 0.85 0.85
Coefficient k
hc
Type-2 EO 1.75 1.75
Superstructure Weight W
u
(tf) 633 474 (minimum)
b) Lateral Strength and Residual Displacement
This chapter follows the checking procedure in the longitudinal direction, and the
formulas and numbers in the brackets, [ ], are for the transverse direction.
Firstly, the strain-stress relationship of the concrete is calculated. The effective length
of transverse reinforcement d for the longitudinal direction is 1.0m [transverse:0.88m] from
Fig-5.3.6, which is the longest centre-to-centre transverse distance between the main-bars
next to the cross-bars. Then, the volumetric ratio of the transverse reinforcement is
52
calculated by Eq(4.4.31).
Longitudinal: Ps =(4x1.986) / (15.0x100.0) =0.00530::; 0.018
[transverse: Ps =(4x1.986) / (15.0x88.0) =0.00602 s 0.018]
From Eq(4.4.25), the strain-stress relationship of the concrete is obtained as shown in Fig-
5.3.7.
53
250 -r------------------------------,
0'cc=222.1
0.80'cc=177.7
200
N
E 150
(J
;;:::
:
"-'
III
III
~ 100
U5
50
I
I
I
---------,---------
I
I
I
I
I
I
I
I
I
I
I
&cc=0.00300 &cu=0.00443
I
I
I
0.0005 0.001 0.0015 0.002 0.0025 0.003 0.0035 0.004 0.0045 0.005
Strain
a) Longitudinal Direction
250,-------------------------------,
ecu=0.00477
&cc=0.00313
I
I
I
---------r---------
O'cc=223.7
50
200
N 150
E
(J
~
III
III
~ 100
0.006 0.005 0.004 0.003
Strain
b) Transverse Direction
0.002 0.001
O-f-----t-------+-----+-..:...-----+------..:'--+------!
o
Figure-5.3.7 Strain-Stress Relationship of Confined Concrete
54
The designed section is checked for Type-1 earthquake. The ultimate strain of the
concrete for Type-1 earthquake is determined as B
eu
= Bee = 0.00300 [transverse: 0.00313],
as seen in Fig-5.3. 7. Here, the ultimate strain of the concrete is assumed to occur at the
outer layer of the main-bars. The relationship between bending moment and curvature at
the base section of the column is obtained following the procedure described in 4.4.2, and
results are seen as follows.
Cracking point : Me =1269.6 tt-rn [2736.7 tf-rn],
tPe = 1.016 X 10-4 m" [4.470 X 10-
5
m"]
Initial yielding point : M
yo
=4373.6 tt-m [7476.5 ttm],
yo = 1.052 X 10-3 m" [4.432 x 10-4 m-
1
]
Reference yielding point: My = 4956.6 tt-m [10868.0 tt-rn],
y= 1.192 x 10-3 m' [6.450 x 10-4 m"]
Ultimate limit state : M
u
=4956.6 tt-rn [10868.0 tt-rn],
tPu = 1.417 X 10-
2
m" [3.304 x 10-3 rn"]
The relationship between lateral load and lateral displacement at the superstructure mass
is obtained as follows.
Cracking point
Initial yielding point
Reference yielding point
Ultimate limit state
: Pc =129.7 tf [218.9 tf]
: Pyo = 437.4 tf [597.4 tf], ~ o = 0.0308 m [0.0189 m]
: P
y
=495.7 tf [869.4 tf], ~ =0.0349 m [0.0275 m]
: Pu =495.7 tf [869.4 tf], Ou =0.1697 m [0.0886 m]
Inthis stage, the plastic hinge zone length is obtained by Eq(4.4.14).
L
p
= 0.2 x 10.00 - 0.1 x 2.20
=1.78m > 0.5 0 =0.5 x 2.20 =1.10m, (0.10:s: Lp:S: 0.5D)
:. L
p
= 1.10 m [2.00 m]
Figures-5.3.8 and 5.3.9 show the relationship between bending moment and curvature at
the base of the column, and the lateral strength and lateral displacement at the
superstructure mass, respectively.
55
Figure-5.3.8 Moment-Curvature Relationship at the Base
Section of the Column (for Type-1 earthquake)
600
Y
500
Yo
g 400
.c
-
Cl
C
~
300
-
(/)
~
IV
-
200
III
...J
100
0
0 0.1 0.2
Yo : Initial Yield
Y : Yield
U1 : Ultimate
0.3 0.4
Displacement (m)
Figure-5.3.9 Lateral Strength-Displacement Relationship at
the Centre of the Superstructure (for Type-1 earthquake)
56
The lateral shear strength of the column for Type-1 earthquake is obtained by
Eq(4.4.32) to Eq(4.4.34) inclusive.
Sc =10 x 0.6 xO.845 x 1.327 x 3.3 x 5.00 x 2.03 =225.4 tf [168.0 tf]
s, =(11.916 x 3000 x 2.03)/ (10 x 1.15 x 15.0) =420.7 tf[842.8 tf]
P =Sc + S5 =225.4 +420.7 =646.1 (tf) [1010.8 (tf)]
The column should fail by flexure in both directions in the event of Type-1 earthquake
because of following conditions.
P =495.7 tf < P =646.1 tf [P
u
=869.4 tf < P
s
=1010.8 tf]
and P; =129.7 tf < P =495.7 tf [Pc =218.9 tf < Po =869.4 tf]
The allowable ductility factor /la for Type-1 earthquake is obtained by Eqs(4.4.12).
/la = 1 + (0.1697 - 0.0349 ) / ( 3.0 x 0.0349 ) =2.29 [1.74]
From the design lateral seismic coefficient k
hc
seen in Table-5.3.5, the equivalent lateral
seismic coefficient k
he
is obtained by Eq(4.4.10).
k
hc
= 0.85 [0.85]
k
he
=0.45 [0.54] ( ~ O . 4 C
z
=0.4, OK)
The equivalent weight W for Type-t earthquake is determined by Eq(4.4.5).
w=633.0 +0.5 x 346.2 =806.1 tf [647.1 tf]
where, C
p
is equal to 0.5 because of the failure mode of the column. The lateral strength of
the column of P1 is determined and compared with the design lateral force as follows.
P
a
=P
u
=495.7 tf> k
he
W =0.45 x 806.1 =362.7 tf
[P
a
=P =869.4 tf> k
he
W= 0.54 x 647.1 =349.4 tf
-------- O.K.
-------- O.K.]
It is confirmed that the lateral strength of the column of P1 is greater than design lateral
force of Type-1 earthquake, for the longitudinal direction.
57
The residual displacement of the column for Type-t earthquake is obtained by Eq(4.4.7)
and Eq(4.4.8). and checked by Eq(4.4.6).
= ~ { ( 0 . 8 5 X 806.1)2 + 1} =1.46 [07<1 0]
j.1R 2 495.7 . .
OR = 0.6 ( 1.46 - 1 ) ( 1 - 0 ) 0.0349 = 0.01 m [Om]
OR= 0.01 m < ORa =h 1100 =10/100 =0.1 m ------- O.K.
Because the response ductility factor for the transverse direction is less than 1.0, the
response will be elastic and therefore there will be no residual displacement in the
transverse direction, for Type-t earthquake. The residual displacement for Type-1
earthquake for the longitudinal direction is less than the limit.
The same step is followed for the Type-2 earthquake. The ultimate strain of the
concrete for Type-2 earthquake is determined as ccu = 0.00443 [0.00477]. as seen in Fig-
5.3.7. The relationship between the bending moment and the curvature at the base section
of the column is obtained as follows.
Cracking point Me = 1269.6 tt-m[2736.7 tt-rn],
c = 1.016 x 10-4 m' [4.470 x 10-
5
m"]
Initial yielding point : M
yo
= 4373.6 tf-rn [7467.5 tt-m],
yo =1.052 x 10-
3
m' [4.432 x 10-4 rn']
Reference yielding point: My = 4956.6 tf-rn [10903.1 tt-rn],
y = 1.192 x 10-
3
rn' [6.471 x 10-4 rn"]
Ultimate limit state : M
u
= 4956.6 tf-rn [10903.1 tt-m],
u = 2.276 x 10-
2
m' [5.361 x 10-
3
m-
1
]
The relationship between the lateral load and lateral displacement at superstructure mass
is obtained as follows.
Cracking point : Pc =129.7 tf [218.9 tt],
Initial yielding point : P
yo
=437.4 tf [597.4 tt], bYo =0.0308 m [0.0189 m]
Reference yielding point : P; = 495.7 tf [872.2 tf], bY = 0.0349 m [0.0276 m]
Ultimate limit state P; = 495.7 tf [872.2 tt], Ou = 0.2590 m [0.1360 m]
58
C : Crack
Yo: Initial yied
Y:Yield
U2 : Ultimate (Type-2)
6000
Y
5000
Yo
4000
E
~
C
3000
(])
E
0
~
2000
C
1000
0
0 0.005 0.01 0.015 0.02 0.025
Curvature (11m)
Figure-5.3.10 Moment-Curvature Relationship at the
base of the column (for Type-2 earthquake)
Yo : Initial Yield
Y : Yield
U2 : Ultimate
600
Y
500
T
~ Yo
g 400
.c
OJ
c
(])
300
'-
U5
(ii
....
<ll
<il
200
...J
100
0
0 0.1 0.2 .
Displacement (m)
0.3 0.4
Figure-5.3.11 Lateral Strength-Displacement Relationship at
the Centre of Superstructure Mass (Type-2 earthquake)
59
The lateral shear strength of the column for Type-2 earthquake is obtained from
Eq(4A.32) to Eq(4A.34) inclusive.
Sc =10 x 0.8 xO.845 x 1.327 x 3.3 x 5.00 x 2.03 =300.5 tf [224.0 tf]
Ss =(11.916 x 3000 x 2.03)/ (10 x 1.15 x 15.0) =420.7 tf [842.8 tf]
P, =Sc + Ss = 225.4 + 420.7 =721.2 tf [1066.8 tf]
The column should fail by flexure in both directions in the event of Type-2 earthquake
because of following conditions.
P
u
=495.7 tf < P =721.2 tf [P
u
=872.2 tf < P =1066.8 tf]
and P; =129.7 tf < P =495.7 tf [Pc =218.9 tf < P =872.2 tf]
Then, the allowable ductility factor fJa for Type-2 earthquake is obtained by Eq(4A.12).
fJa = 1 + (0.2590 - 0.0349 ) / ( 1.5 x 0.0349 ) =5.28 [3.62]
From the design lateral seismic coefficient k
hc
seen in Table-5.3.5, the equivalent lateral
seismic coefficient k
he
is obtained by Eq(4A.1 0).
; = 1.75 [1.75]
k
he
= 0.57 [0.70] (:2: 004 C
z
= 004, OK)
The equivalent weight Wfor Type-2 earthquake is determined by Eq(4A.5).
-W= 633.0 + 0.5 x 346.2 = 806.1 tf [647.1 tf]
The lateral strength of the column of P1 is determined and compared with the design lateral
force as follows.
P
a
= P = 495.7 tf> k
he
W= 0.57 x 806.1 = 459.5 tf
[P
a
= P
u
= 872.2 tf > k
he
W= 0.70 x 647.1 = 453.0 tf
------- O.K.
-------- O.K.]
It is confirmed that the lateral strength of the column of P1 is greater than the design lateral
force of Type-2 earthquake, for both the longitudinal and transverse directions.
60
The residual displacement of the column for Type-2 earthquake is obtained by
Eq(4.4.7) and Eq(4.4.8), and checked by Eq(4.4.6).
: +1}=4.55 [1.34]
j.1R 2 495.7
OR: 0.6 ( 4.55 - 1 ) ( 1 - 0) 0.0349: 0.074 m [0.006m]
OR: 0.074 m < ORa: h /100: 10/100: 0.1 m ------- O.K.
[OR: 0.006 m < ORa: h / 100: 12.5/ 100 : 0.125 m -------- O.K.]
The residual displacement for Type-2 earthquake is less than the limit for both the
longitudinal and transverse directions.
c) Checking Result of Lateral Strength Method
Table-5.3.6 summarises the checking procedure of the Lateral Strength Method. The
designed section of the column of P1 seen in Fig-5.3.6 is confirmed to be safe against both
Type-1 and Type-2 earthquake.
Table-5.3.6 Checking Result of the Designed Section of the column of P1
Longitudinal Transverse
Type-1 EO Type-2 EO Type-1 EO Type-2 EO
Lateral KheW (tf) 362.7 459.5 349.4 453.0
Strength Pa (tf) 495.7 495.7 869.4 872.2
Judgement P
a
> KheW P
a
> KheW r.> KheW r.> KheW
OK OK OK OK
Residual OR (m)
0.010 0.074 0 0.006
Displacement
ORa (m)
0.100 0.100 0.125 0.125
Judgement
OR < ORa OR < ORa OR < ORa OR < ORa
OK OK OK OK
5.3.4 Modification of the Section
5.3.4.1 General
The rectangular section designed in 5.3.3 was modified to a square section seen in
Fig-5.3.12, following the same procedure. Columns with relatively small sections and
61
subjected mainly to flexure like this example are determined by the Lateral Strength Method
because of its much greater design seismicity which directly increases the flexure on the
column. Therefore, this square section is checked by only the Lateral Strength Method
here. The longitudinal reinforcement is doubly symmetrical; the design strength of the
concrete is increased from 210kgf/cm
2
(20.6MPa) to 240kgf/cm
2
(23.5MPa) and the yield
strength of the reinforcement is also increased, from 3000kgf/cm
2
(295MPa) to 3500kgf/cm
2
(345MPa). The diameter of the main-bars is 41mm and the diameter of the transverse
reinforcement is 19mm with 150mm vertical spacing.
3 000
130 16 160
14@150= 2 100
160 68 no
I I
~ 7
~
~
~
J2
J2
r--
~
C>
C>
C>
0
('lJ
~
~
~
nb 620 750 750 670 1l"1n
3 000
Figure-5.3.12 Square Section for the Column of P1 [unitmm]
5.3.4.2 Design Lateral Seismic Coefficient and Superstructure Weight
Table-5.3.7 shows the values of design seismic coefficient, superstructure weight and
height of the superstructure mass, used to design the prototype column with the square
section. The largest design seismic coefficients for the intermediate soil condition are used
for both Type-1 and Type-2 earthquakes, same as the previous design example of the
rectangular section, but without any calculation of the natural period. The height from the
base of the column to the superstructure mass h, is determined as 12.5m which is for the
transverse direction seen in Fig-5.3.2. Because the aspect ratio is more than 4 this column
should fail by flexure rather than shear. The superstructure weight W
u=633(tf)
is taken as
62
the greater value from Table-5.3.5.
Table-5.3.7 Design Lateral Seismic Coefficients for Lateral Strength Method
and Superstructure Weight Supported by Prototype Column
Design Lateral Seismic Type-1 EO 0.85
Coefficient k
h
Type-2 EO 1.75
Superstructure Weight Wu (tf)
633
Height of Lateral Loading Point h (m)
12.5
5.3.4.3 Volumetric Ratio of Transverse Reinforcement
The effective length of transverse reinforcement d, for the longitudinal direction is
assumed to be equal to 0.75m from Fig-5.3.12. The volumetric ratio of the transverse
reinforcement Ps is calculated by Eq{4.4.31).
Ps =(4x2.865) / (15.0x75) =0.0102:s; 0.018
From Eq{4.4.25), the strain-stress of the concrete is obtained as shown in Fig-5.3.13.
300 -r---------------------------,
o-cc=267.1
0.008 0.007 0.006
s cu =0.00691
0.005 O. 04
Strain
0.003
s cc=0.00396
I
I
I
- - - - - - - - ~ - - - - - - - - - - - -
I
I
0.002 0.001
0.80-cc =213.7
o +,---+-----+-----+---.li----+----+----..I..+----l
o
50
250
....... 200
C'\I
E
o
;;:::
Ol
150 ~
.......
Ul
Ul
(])
L..
.....
(/)
100
Figure-5.3.13 Strain-Stress Relationship of the Concrete (Square section)
63
5.3.4.4 Type-1 Earthquake
The ultimate strain of the concrete for Type-1 earthquake is determined as CcU = Gee =
0.00396, as seen in Fig-5.3.13, and is assumed to occur at the main-bars in compression.
The relationship between the bending moment and the curvature at the base of the column
is obtained as follows.
Cracking point : Me = 1477.9 tt-rn , rPe = 7.7093 x 10-5 rn"
Initial yielding point : Myo=4487.7 tt-rn I rjJya =8.4522 X 10-4 rn"
Reference yielding point : My =5624.2 tt-m , rjJy =1.0593 x 10-3 rn"
Ultimate limit state : M
u
=5624.2 tt-rn I rPu = 1.0719 x 10-
2
rn"
The relationship between the lateral load and lateral displacement at the superstructure
mass is obtained as follows.
Cracking point : P
e
= 118.2 tf ,
Initial yielding point : Pyo=359.0 tf , bYo =0.0370 m
Reference yielding point : P, = 449.9 tf , bY =0.0464 m
Ultimate limit state : P =449.9 tf I ~ = 0.2204 m
The plastic hinge zone length is obtained by Eq(4.4.14).
L
p
= 0.2 x 12.50 - 0.1 x 3.00
= 2.20m > 0.50 = 1.5m
.. L
p
=1.50 m
Figures-5.3.14 and 5.3.15 shows the relationships between bending moment and curvature
at the base of the column, and lateral strength and lateral displacement at the centre of the
mass of the superstructure, respectively.
64
C : Crack
Yo: Initial yied
Y: Yield
U1 : Ultimate (Type-1)
7000
6000
Y
5000
E
Yo
~
4000
......,
.....
c
Q)
E
3000
0
::2
2000
C
1000
0
0 0.005 0.01 0.015 0.02 0.025
Curvature (11m)
Figure-5.3.14 Moment-Curvature Relationship at the base of
the column (Square section, for Type-1 earthquake)
y
Yo
600
500
g 400
.c
.....
Q')
c
Q)
300
....
U5
Cii
....
2
ro
200
..J
100
0
0 0.1 0.2
Yo : Initial Yield
Y : Yield
U1 : Ultimate
0.3 0.4
Displacement (m)
Figure-5.3.15 Lateral Strength-Displacement Relationship at
the Centre of Superstructure Mass
(Square section, for Type-1 earthquake)
65
The lateral shear strength of the column for Type-1 earthquake is obtained from
Eq(4.4.32) to Eq(4.4.34) inclusive.
Sc = 10 x 0.6 xO.720 x 1.222 x 3.5 x 3.00 x 2.87 = 159.1 tf
Ss = (14.325 x 3500 x 2.87) I (10 x 1.15 x 15.0) = 834.2 tf
P = Sc + Ss = 159.1 + 834.2 = 993.3 (tf)
The column will fail by flexure in the event of Type-1 earthquake because of following
conditions.
P = 449.9 tf < P, = 993.3 tf
and Pc = 118.2 tf < Po = 449.9 tf
Then, the allowable ductility factor fla for Type-1 earthquake is obtained by Eq(4.4.12).
fla = 1 + (0.2204 - 0.0464) I ( 3.0 x 0.0464) =2.25
The equivalent lateral seismic coefficient k
he
is obtained as follows.
k
hc
= 0.85
k
he
= 0.46 ( ~ 0 . 4 C
z
= 0.4, OK)
The equivalent weight Wfor Type-1 earthquake is determined by Eq(4.4.5).
W= 633 + 0.5 x 308.8 = 787.4 tf
Where, C
p
is equal to 0.5 because of the failure mode of the column. The lateral strength
of the prototype column is determined and compared with the design lateral force as
follows.
P = P = 449.9 tf > k
he
W= 0.46 x 787.4-= 362.2 tf ------- O.K.
It is confirmed that the lateral strength of the prototype column is greater than the design
lateral force of Type-1 earthquake, for the longitudinal direction.
The residual displacement of the column for Type-1 earthquake is then checked. The
66
residual displacement is obtained by Eq(4.4.7) and Eq(4.4.8), and checked by Eq(4.4.6).
: ~ { ( 0 . 8 5 X 787.4)2 +1} =1.61
j.1R 2 449.9
OR: 0.6 ( 1.61 - 1 ) ( 1 - 0 ) 0.0464 : 0.017 m
OR: 0.017 m < ORa: h 1100: 12.51100: 0.125 m ---- O.K.
The Residual displacement for Type-1 earthquake for the longitudinal direction is less than
the limit.
5.3.4.5 Type-2 earthquake
The ultimate strain of the concrete for Type-2 earthquake is determined as E
cu
:
0.00691, as seen in Fig-5.3.13. The relationship between the bending moment and the
curvature at the base of the column is obtained as follows.
Cracking point : Me : 1477.9 tt-rn , !/Je: 7.709 X 10-5 11m
Initial yielding point : M
yo
: 4487.7 tf-m , yo: 8.452 x 10-41/m
Reference yielding point : My : 5593.8 tf-rn , I/Jy: 1.054 x 10-3 11m
Ultimate limit state : M
u
: 5593.8 tf-rn , !/Ju : 2.048 x 10-
2
11m
The relationship between the lateral load and the lateral displacement at the centre of the
mass of the superstructure is obtained as follows.
Cracking point : Pc : 118.2 tf ,
Initial yielding point : Pya: 359.0 tf , bya: 0.0370 m
Reference yielding point: P
y
: 447.5 tf , by: 0.0461 m
Ultimate limit state : P
u
=447.5 tf , b
u
=0.3921 m
67
0.025 0.02
C : Crack
Yo : Initial yied
Y:Yield
U2 : Ultimate (Type-2)
0.015
Curvature (11m)
Figure-5.3.16 Moment-Curvature Relationship at the base of
the column (Square section, for Type-2 earthquake)
7000
6000
Y
5000
Y
a
~ 4000
.....
c
Q.J
E
3000
0
~
2000
C
1000
0
0 0.005 0.01
Yo : Initial Yield
Y : Yield
U2 : Ultimate
600
500
Y
400
1
IY
a
..c:
.....
OJ
c
Q.J
...
300 .....
(fJ
~
Q.J
.....
200
co
...J
100
0
0 0.1 0.2 0.3 0.4
Displacement (rn)
Figure-5.3.17 Lateral Strength-Displacement Relationship at
the Centre of Superstructure Mass
(Square section, for Type-2 earthquake)
68
The lateral shear strength of the column for Type-2 earthquake is obtained from
Eq(4.4.32) to Eq(4.4.34) inclusive.
s; = 10 x 0.8 xO.720 x 1.222 x 3.5 x 3.00 x 2.87 = 212.1 tf
S; = (14.325 x 3500 x 2.87)/ (10 x 1.15 x 15.0) = 834.2 tf
P= Sc + Ss = 212.1 + 834.2 = 1046.3 (tf)
The column will fail by flexure in the event of Type-2 earthquake because of following
conditions.
P
u
= 447.5 tf < P = 1046.3 tf
and P
c=212.1tf<Pu=447.5tf
Then, the allowable ductility factor J.1a for Type-2 earthquake is obtained by Eq(4.4.12).
J.1a = 1 + (0.3921 - 0.0461) / ( 1.5 x 0.0461) =6.00
The equivalent lateral seismic coefficient k
he
is obtained by Eq(4.4.10).
k
hc=1.75
k
he
= 0.53 (;::: 0.4 C
z
= 0.4, OK)
The equivalent weight Wfor Type-2 earthquake is determined by Eq(4.4.5).
W = 633 + 0.5 x 308.8 = 787.4 tf
The lateral strength of the column of P1 is determined and compared with the design lateral
force as follows.
P, = P = 447.5 tf > k
he
W= 0.53 x 787.4 = 417.3 tf -------- O.K.
It is confirmed that the lateral strength of the column of P1 is greater than the design lateral
force of Type-2 earthquake, for the longitudinal direction.
The residual displacement of the column for Type-2 earthquake is
69
= ~ { ( 1 . 7 5 X 787.4)2 +1} =5.24
j.1R 2 447.5
OR =0.6 ( 5.24 - 1 ) ( 1 - 0 ) 0.0461 =0.117 m
OR= 0.117 m < ORa =h /100 =12.5/100 =0.125 m -- O.K.
Residual displacement for Type-2 earthquake is less than the limit.
5.3.5 Design of Model Specimen
The reinforced concrete column with 3mx3m square section designed in 5.3.4 is
scaled down for the test specimen. The specimen is with 550mmx550mm square section
and 3.5m high as seen in Fig-5.3.18.
This dimension is the maximum size possible which can be tested by the experimental
facilities in the laboratory of the University of Canterbury. The specimen is designed to
follow the scale ratio equal to 5.45, which is the ratio 3m to 550mm, for as many details as
possible, in order to avoid the scale effect in the experiment. Generally, scale effects can
lead to additional column rotation at the base of the column due to the yield penetration of
the main-bars. In addition, the lateral depth of the column can affect the available concrete
shear stress. The diameters of the reinforcement is 10mm and 6mm, for the main-bars and
transverse reinforcement, respectively. These bars were the minimum size of bars
available in New Zealand when the experiment was planned. The difference of the yield
strength of the reinforcement from that used for the prototype column is adjusted by the
total area of the bars. The concrete is designed to have the same strength as the
prototype, with the 6mm-maximum aggregate, the smallest aggregate available in New
Zealand. Table-5.3.8 shows the comparison of major factors between the prototype column
and the specimen.
70
550
152 190 , 152
28
28
\
~
~
6 6
<::>
'"
'"
,
c
13(0)38= 494
A-A
Figure-5.3.18 Reinforcement Arrangement of the Specimen [unitmm]
71
Table-5.3.8 Comparison Between Prototype and Specimen
Prototype Specimen Ratio Comments
Section (mm)
3000*3000 550*550 5.45
Cover (from cl of main-bar) (mm) 130 28 4.64
Cover (pure) (mm)
71.5 11 6.47
Height of loading point (mm) 12500 2250 5.56
Concrete
Design strength (MPa) 23.5 23.5 1.00
f'ck
Max size of aggregate 25 6 4.17
(mm)
Main- Yield strength (MPa) 345 300
bars Diameter (mm) 41 10
Sectional area (rnrrr) 1340 78.5 For one
bar
Total number (mm) 76 52
Steel ratio (%) al 1.132 1.174 0.964
As/lAg
Trans- Yield strength (MPa) 345 300
verse Diameter (mm) 19 6
Bars Sectional area (rnm") 286.5 28.3 A
h
Vertical c-to-c spacing (mm) 150 50 s
Effective length (mm) 750 190 d
Steel ratio (%) OJ
1.019 1.035 0.984
AtI(sd)
Axial Axial load (kN) 9239 311 N
Load Stress (MPa) 1.027 1.027 1.00 N/A
g
Stress I Capacity (%) 4.36 4.36 1.00
N/Agf'ck
a) The yield strength of the reinforcement IS considered,
5.4 Design of Testing Equipment
5.4.1 Modification of the Rig
The L-shaped loading rigs, seen in Fig-5.4.1. were originally designed by Wong et al.
[20] for bi-directional cyclic loading tests on some circular short column specimens. The
vertical members of the rigs were elongated by approximately 1.5m for these tests in order
72
to raise the loading point from 770mm to 2250mm above the base of the column. A
350mm thick reinforced concrete spacer was used between the base-block and each of the
rigs because of the longer horizontal jacks used to apply larger displacements.
20mm Side at
ll---..,If--+---U
2370
r-
2
,
a) Elevation
20mm Sliffenfl"s
i/o
9
J.
T
'"
r
" ..
1
1
In 60mm PlatE'
---- --._--_.-._._.
o -500mm
Scale
b) Plan
Figure-5.4.1 Loading Rig (Wong et al. [20])
5.4.2 Check of Testing Equipment
Prior to the check of the testing equipment, the possible lateral strength and yield
displacement of the test specimen were calculated using a spread sheet written for the
design of the prototype column using the Japanese specification [1]. described in 4.4.
The calculated values are
73
The lateral strength of the column at the loading point = 159.7 kN
The reference yield displacement at the loading point = 8.14 mm
The possible maximum lateral strength and the reference yield displacement can be
assumed using the flexural overstrength factor 1.47 derived from the New Zealand
Specification [21].
Possible maximum lateral strength = 159.7 x 1.47 = 234.8 kN
Possible reference yield displacement = 8.14 x 1.47 = 12.0 mm
5.4.2.1 Strength of the Loading Rig
The horizontal load capacity of the loading rig at the loading point, 2250mm from the
base of the column, was calculated and compared to the possible maximum lateral load.
The horizontal load capacity was obtained from the moment capacity of the loading rig at
the centre of the elbow, which is the joint of the lateral member and the vertical member,
and at the joint between the original vertical member and the newly welded vertical
extension piece.
a) Moment Capacity at the elbow
Assuming the whole section of two 381 mm channels and the two 20mm
x400mm
skin
plates are effective against the flexure at the centre of the elbow, the second moment of
area I is obtained as 884.6x10
6
mm". The moment capacity at the elbow is obtained as
follows, defining it as the moment when the edge of the section yields,
fI
M=2:...- x10-
6=
1261 kNm
y
where, f
y
is the yield stress of the steel =300MPa and y is the distance from the centre of
the section to the outer fiber =21 0.5mm. The maximum horizontal load at the loading point
L
max
is calculated from M and the height of the loading point from the centre of the elbow h
= 2576mm, as
M
L
max
=-10
3
= 489 kN 234.8 kN
h
The moment capacity at the elbow is more than twice the requirement.
74
b) Moment Capacity at the Joint
A 381mm
x554mmx20mm
plate was welded on the top of the vertical member then four
381mm channels were combined and welded on the plate to extend the vertical member.
Moreover, additional 20mm thick plates covered the holes below the joint. Assuming four
381mm channels are effective at the joint, the second moment of area " the moment
capacity M and the maximum horizontal load at the loading point L
max
are obtained as
follows.
fI
M=Lx 10-
6
=958 kNm
y
L
max
=M10
3
= 778 kN 234.8 kN
h
Where, h = 1231mm
y= 190.5mm
The moment capacity at the joint section is more than twice the requirement.
c) Connection Bolt Strength
The number of the bolts connecting the loading rig to the base-block is nine, five at the
top and four at the bottom, as seen in Fig-5.4.1. The centre-to-centre vertical distance of
the bars is 530mm and the distance between the top bars and the horizontal loading point
is 2361mm. The diameter of the bolts was 24mm, and the yield strength f
y
, using a sample,
was 800MPa.
When the top bolts are yielding in tension, the corresponding lateral load at the loading
point is calculated as follows assuming the centre of the compression is at the bottom bolts.
L= (24
2
TC / 4)x 800 x 5 x 530 xi 0-
3
= 332 kN 234.8 kN
(2361+530)
When the bottom bolts are yielding in tension and the centre of the compression is
assumed at the top bolts, the corresponding lateral load at the loading point is
75
L =(24
2;r/4)x800x4
x530x10-
3
=325 kN 234.8 kN
2361
The tensile strength capacity of the bolts was about twice greater the requirement, for both
the top bolts and the bottom bolts.
5.4.2.2 Horizontal Hydraulic Jacks
The available travel of the horizontal hydraulic jack used for the tests was 400mm,
namely 200mm for each direction from the centre. Therefore, the applicable displacement
ductility is
200mm / 12mm= 16.7
This value is great enough considering the design available ductility factor of the column
,lla=6.0 calculated in 5.4.4.5. The maximum loading strength of the jack was 300kN, which
is about 30% greater than the possible maximum load 234.8 kN.
5.5 Measurements
5.5.1 Lateral Displacement Measurement
Three potentiometers in each loading direction measured the displacement of the test
specimen at the loading point, at three-quarters and at the half height of the column above
the base. The potentiometers were attached to an vertical plate having the same height as
the test column and fixed to the counter weight basket at the bottom, as seen in Fig-5.5.1.
Each potentiometer was connected to the specimen through a string tied on an anchor at
the measured point on the test specimen.
76
String
,r-
Potentio-
metres
Test
Column
Co unt er Wei 9ht Sa sket
Figure-5.5.1 Horizontal Displacement Measurement
5.5.2 Curvature Measurement
Five potentiometers are mounted on each surface of the test specimen, at 80mm,
230mm, 430mm, 630mm and 830mm from the base of the column respectively, on the
vertical centre line of the face (see Fig-5.5.2). The curvature is calculated from the
contraction or expansion measured by the potentiometers on the opposite sides of the test
specimen in a plane parallel to the loading direction, by Eq(5.5.1).
(5.5.1)
Where, d
1
and d
2
are the contraction or expansion measured on the opposite sides at the
same level; L
pot
is the horizontal distance between the potentiometers mounted on the
opposite sides; and Xpot is the vertical distance between two potentiometers mounted in the
same face of the test specimen where d, or d
2
is measured (e.g. 80mm, 230mm etc.)
77
Figure-5.5.2 Potentiometers for Curvature Measurement
5.5.3 Strain Measurement
The strain on eight main bars in each test specimen, on the four corner-bars and on
the centre-bars of the four faces, were measured by the strain-gauges. The main bar used
for strain measurement had four strain-gauges attached at 50mm below the base of the
column, and at 20mm, 220mm and 470mm above the base. The strain on two layers of the
transverse reinforcement were also measured, at 50mm and 250mm from the base. Each
layer of transverse reinforcement measured had four strain-gauges on the hoop and two
strain-gauges on the cross-bars. The location of the strain-gauges are seen in Fig-5.5.3,
and Fig-5.5.4
78
II
II
~
~ ~ ~
~
~
~
~ ~
0
L()
N
0
0
N
0
C'?
~ ~
Figure-5.5.3 Location of Strain-Gauges on Vertical Bars [unitmm]
18J:Straingauge
-a-a -.;:r
u v v v
v v ~
b
P (
0 (
-=
p
C
P
c
b
c
0
p
C
P
c
10 c
- -
~ ~
n n ~
Figure-5.5.4 Location of Strain-Gauges on Hoop and Cross-Bars
5.6 Control and Data Acquisition
The axial load on the test specimen was automatically stabilised by the DARTEC, and
the lateral hydraulic jacks were manually controlled. The test data was collected by five
data boxes, each of which has 16 channels, then recorded on a personal computer using
the software developed in the University of Canterbury.
79
a) Hydraulic Jack Controller
b) Personal Computer For Data Logging
Figure-5.5.5 Data Acquisition System
80
Chapter VI
Test
6.1 Construction of Test Specimens
Two sets of moulds were made for five test specimens. The order in which the
columns and column bases were constructed is as follows. Two sets of the reinforcement
for the base-block were assembled, as seen in Fig-6.1.1, placed in the form and the
concrete poured, Fig-6.1.2. Then the transverse reinforcement (Fig-6.1.3) was assembled
for the two columns and one new base-block was also assembled. Concrete was poured
to complete the concreting of two columns and one base-block. A concrete pump was
used to pour the concrete into the columns because the total height of the specimens was
3.5m from the floor. The transverse reinforcement was assembled for the third column,
and the fourth and fifth base-blocks were assembled, then a third concrete pour was
completed. Here, as mentioned in 5.1, a wrong mixture was accidentally used for the third
concrete pour. It was found in the seventh-day-cylinder test that the compressive strength
of the third concrete pour was less than half of the design strength, so then this concrete
was removed from the fourth and fifth base-blocks by a concrete breaker. Because the
reinforcement mesh of the column was too fine, however, the low-strength concrete in the
third column was impossible to remove. The third specimen with the extremely weak
concrete was tested as Specimen-5.
After the concrete was poured into a base-block, some retarder was applied on the surface
of the concrete at the construction joint between the column and the base-block, and then
the concrete surface at the construction joint was coarsened by a steel brush a few days
later.
81
Figure-6.1.1 Reinforcement of a Base-block
Figure-6.1 .2 Concrete Pour for Base-block
82
Figure-6. 1.3 Reinforcement of Column
6.2 Propert ies of Materials
6.2.1 Reinforcing Steel
6mm and 10mm diameter deformed bars were used for the transverse reinforcement
and the main vertical bars, respectively. The design strength was 300MPa for both types of
reinforcing bars. 20mm diameter high-strength deformed reinforcing bars, with a design
strength of 430MPa, were used in the base-blocks, but were not tested.
83
The tensile strength of the reinforcing bars were tested using four samples of 6mm
deformed bars and three samples of 10mm deformed bars. Fig-6.2.1 and Fig-6.2.2 show
the stress-strain relationship of the test samples of the 6mm deformed bars and the 10mm
deformed bars, respectively. Table-6.2.1 shows the test results.
84
600 ...-----:-------:------;------:-------:-----,
200
--06-1
-06-2
06-3
- 06-4
-- -- -- - - - - - -- - - - - - -- - - ~ - - - - - - - - - - - - - - - - - -- - - - - ~ - - - - - - -- -- - - -- -- - - - - - - -- ---- - - --- - --~ ----- .-
, ,
, ,
- - - - - - - - -- - - - - - _. -- -., - - - - -- - - -- - - ----_.- --- - - --- - --- - -.- - - -- -- - -, - _. - - - - - --- --. - - --- - - - -,- - _. - - - - - -- - - _. - - - - - --- . - --- -- _. -- --
400
500
..-.
ro
o,
:2:
(j) 300
en
~
U5
0.12 0.1 0.08 0.06
Strain
0.04 0.02
o -11-, -----;-----+------+-----+------;-------;
o
Figure-6.2.1 Strain Stress Relationship of 6mm Deformed Bars
:
: :
----
~
-
"T'-:::--
~ - - -
-. ...
AI
v-:
----------
.-
---
:
010-1
----------------
-
-010-2
----_. ---- - - ----
. .
-
010-3
, ,
500
450
400
350
..-.
ro
300
o,
:2:
.....,
250
en
en
<1l
...
200 .....
(f)
150
100
50
o
o
0.05 0.1 0.15 0.2
Strain
0.25 0.3 0.35 0.4
Figure-6.2.2 Strain Stress Relationship of 10mm Deformed Bars
85
Table-6.2.1 Test Results of Reinforcing Bars
Samples
f
y
E
s
f
su
li
su lis'
Re-bar size No.
(MPa) (MPa) (MPa)
1 279.4 203100 488.1 0.0814 0.1059
2 272.3 263900 481.0 0.0901 0.0987
6mm 3 300.6 194200 491.6 0.0703 0.1063
4 290.0 187300 481.0 0.0703 0.0858
Ave 285.6 212100 485.4 0.0777 0.0992
1 306.9 254900 438.0 0.1766 0.2389
10mm 2 299.2 174000 440.5 0.1997 0.3747
--
3 311.9 181400 438.0 0.1997 0.2846
Ave 306.0 203400 438.8 0.1920 0.2994
-- --
Where; f
y
is the yield stress; E, the Modulus of Elasticity; f
su
is peak stress; li
su
is the strain
at peak stress; and cs' is the fracture strain.
As seen in Fig-6.2.1 and Fig-6.2.2, 6mm bars do not have a yield plateau and are more
brittle than the 10mm bars.
6.2.2 Concrete
The concrete strength was obtained by the standard compression test with 100mm-
diameter test cylinders on 7
th
day, zs" day and the day of the experiment. The mixture of
the concrete was specially designed for this project using the aggregate with a maximum
diameter of 6mm, aiming for a designed strength of 23.5MPa in 28 days. The concrete mix
proportion is shown in Table-6.2.2 and the average concrete compressive strength is
shown in Table-6.2.3.
Table-6.2.2 Concrete Mix Proportion
Material Quantity (kq/rn")
Cement 310 [278]
Coarse aggregate (6mm Max) 1077
Fine aggregate 881
Water 200
[ ] IS for the first batch
86
Table-6.2.3 Concrete Compressive Strength
Average Compressive Strength (MPa)
Batch No. r day 28
U
' day Day of Test
Specimen- Base-block 1 12.9 16.6 19.6
1 Column 2 19.5 26.6 30.7
Specimen- Base-block 1 12.9 16.6 19.6
2 Column 2 19.5 26.6 30.7
Specimen-
Base-block 4 18.7 26.1 29.4
3 Column 5 20.6 27.0 27.0
Specimen- Base-block 4 18.7 26.1 29.4
4 Column 5 20.6 27.0 29.4
Specimen- Base-block 2 19.5 26.6 -
5 Column 3 (Wrong Mix) 5.9
-
10.9
Because the strength of the first batch of concrete was lower than the design strength, as
seen in Table-6.2.3, 30kg/m
3
of the cement was added for the second batch. It was
confirmed later that the concrete of NO.3 batch had about 30% less cement than the design
quantity.
6.3 Test Results
6.3.1 P-oEffect of Tests
Sato et al. [22] tested some reinforced concrete columns at the University of
Canterbury using the same experimental procedure as that used for this project, and
described the P-Oeffect of the tests. From Fig-6.3.1, the P-O effect due to the loading
arrangement in this series of tests is calculated as follows.
therefore,
where,
LJ'/cose a
LJ bcose
LJ'= ~ L J
b
a > 750 + 50::: 800mm
b::: 3500 + 50 + 175::: 3725mm
87
(6.3.1)
(6.3.2)
P ,1'= P 800 ,1 = 0.215L1P
e e 3725 e
(6.3.3)
b
Figure-6.3.1 P-5 effect of the tests (Sato et al. [22])
6.3.2 Theoretical Flexural Strength of the Column
6.3.2.1 Ideal Flexural Strength of the Test Specimens
The ideal flexural strength of the test specimens were calculated using the actual
properties of the materials shown in 6.2, in order to determine the reference yield
displacement of each test specimen. The ideal flexural strength is calculated based on the
following assumptions used in the NZ3101(1995) [21]. The maximum compressive strain
of concrete is 0.003; the stress in the reinforcement is E
s
times the steel strain and the
maximum and minimum stress is equal to +f
y
and -f
y
respectively; the tensile strength of
the concrete is neglected; and the concrete compressive stress is represented by a
rectangular distribution with the stress equal to 0.85 Fe and a length equal to 0.85c. E
s
is
Young's modulus of elasticity of the reinforcement; Fe is the specified compressive strength
of concrete; and c is the distance from the extreme compression fibre to the neutral axis.
The calculated values for the ideal flexural strength for the test specimens are shown in
Table-6.3.1.
88
Table-6.3.1 Ideal Flexural Strength M
n
of Test Specimens
M
n
(kNm)
No. of specimen
Without P-5 With P-5
1 382.6 381.8
2 382.6 381.8
3 378.8 378.0
4 381.3 380.5
5 343.9 343.1
6.3.2.2 Moment-Curvature Analysis
The flexural strength of the test specimens are also estimated using the moment-
curvature analysis method introduced by Mander et al. [23], for a comparison. This method
is briefly summarised below.
A horizontal plane section is assumed at the base of the analysed column and the
section is divided into thin fibres perpendicular to the direction of the bending moment
applied to the column. Here, the section is divided into 56 fibres which consist of three for
the cover concrete and 50 for the rest of the section. The strain at the centre of each fibre
is assumed to be in proportion to the distance from the neutral axis of the section. The
moment resistance at the base section of the column for a given curvature is obtained by
adjusting the location of the neutral axis in order to equilibrate the applied axial load and
the generated axial resistance of the section.
The stress-strain relationship of reinforcing steel can be modelled as consisting of
three regions.
Elastic region (0 :S &s:S By) : f
s
= &sE
s
LOders Plateau region (&y s &s s &sh) : f
s
= f
y
Work hardening region (&sh:S; &5:S &SU): t, = f
5U
+(f
y
- f
su
) ( &su - &s )P
&su - &sh
89
(6.3.4)
(6.3.5)
(6.3.6)
where,
10g( f
su
- f
sh1
]
f
su
- f
y
P =--.+------'--7-
10J S su - S sh1 J
'SSU -Ssh
(6.3.7)
t, is steel stress; Ss is steel strain; f
su
is the ultimate tensile strength; f
Sh1
is stress
corresponding to SsM; Ssh is strain at the onset of work hardening; Ssu is strain corresponding
to t.: SsM is strain within the work hardening region and P=4.5. The stress-strain
relationship in compression is assumed to be identical in absolute terms.
The compressive strength of unconfined concrete is
when 0;:: Se ;:: -0.004
when Se < -0.004
f =
c
2( se )
-0.002 f'
1 ( Se )2 e
+ -0.002
(6.3.8)
where, f
e
is concrete stress; Se is concrete strain. The compressive strength of confined
concrete is
for f =
c
(6.3.9)
where, Sec = -0.22 [ 1 + 5 ( feclf
e
- 1 )] (6.3.10)
(6.3.11)
fcc is confined concrete compressive strength (positive); Sec is concrete strain
corresponding to the maximum value of fee; and Seu is the ultimate concrete strain. Seu can
be determined by Eq(6.3.12) derived from Paulay and Priestley [17].
Seu = 0.004 + 1.4psfyt8
sm
/ fcc
90
(6.3.12)
where, Ps is the volumetric ratio of transverse reinforcement; fyt is yield strength of
transverse reinforcement; and Gsm is the maximum strain of transverse reinforcement. Ps is
obtained as a ratio of the volume of transverse confining reinforcement to the volume of
concrete core. fcc is given by,
(6.3.13)
where, at is a strength enhancement factor considering the confining effect and a2 is a
reduction factor considering the deviation from the bi-equal confining stress. a1 and a2 are
calculated as
at =1.25[1.8 1+ 7.94 F; -1.6 F; -1]
f; t;
a2 = - -0.8] [f +1
F; F; VF;
(6.3.14)
(6.3.15)
where, F, is the maximum confining lateral stress (positive); and f, is the minimum confining
lateral stress (positive). Mander et al. [23] proposed a "smeared model" to obtain the
confining lateral stress from the confining capacity of the transverse confining
reinforcement and the effectiveness of the confinement considering the effectively confined
area of the core concrete, as seen in Fig-6.3.2.
91
W'
ineffectively confined concrete
at the level of the ties
cover concrete
n, longifucinal rebars
Figure-6.3.2 Effectiveness of Confining Steel on Core Concrete (Mander et al. [23])
The ratio, Ke, of the effective lateral confining pressure, f'le, to the lateral confining pressure,
f'l, is assumed to be equal to the ratio of the effectively confined area of the core concrete
A, and the whole area of the core concrete Ace.
r; =fie / r, =A
e
/ Ace (6.3.16)
For a rectangular column section with the clear spacing w' between longitudinal bars and
the clear spacing s' between sets of transverse reinforcement, K
e
is determined by
Eq(6.3.17) assuming that the arching of the concrete seen in Fig-6.3.2 has a parabolic
shape and the initial tangent slope of 45 degree.
(6.3.17)
where, hx" and hy" are centre-to-centre horizontal distance of an extreme hoop confining the
core concrete in x and y directions respectively (Fig-6.3.2); w;' is a centre-to-centre
horizontal distance between adjacent cross-bars or hoops confining concrete; and Psi is the
longitudinal steel ratio defined as the ratio of the total area of longitudinal bars to the area
of confined concrete. The lateral confining pressure f/ is
92
(6.3.18)
where, A
st
is the total sectional area of confining reinforcement (cross-bars and hoops) in
one direction; and fyt is the yield strength of the confining reinforcement. The effective
confining pressure ('Ie for each direction, x and y, is determined by Eq(6.3.16) to Eq(6.3.18)
inclusive, and the larger value is defined as F
I
, other is defined as ft in Eq(6.3.14) and
Eq(6.3.15). For the section of the test specimen, h,", hy" and w, are measured as seen in
Fig-6.3.3.
h"y
W12 W11 W10
v v v v v v v
v v ~
p
W1
p
~
p
(
p (
W2
p
~
p
(
p
(
p
(
W
W3
p
(
p
(
" ,., ,.,
,., ,., ,., ,.,
n n ,.,
W4 W'J W6
h"x
9
8
7
Figure-6.3.3 Application of Moment-CurvatureAnalysis for the Test Specimen
The calculated moment-curvature relationships and lateral load-displacement
relationships for the specimens are shown in Fig-6.3.4 to Fig-6.3.7. Table-6.3.2 shows the
ideal lateral strength of the specimens at the loading point obtained by the moment-
curvature analysis and NZ3101 [21]. derived from Table-6.3.1.
93
250 , . . . - - - - - - - - - - - - - : - - - - - : - - - ~ - - _ : _ _ - - : _ _ - _ _ _ : _ - - _ ,
200 - -- . - . ~ .. __ .- - .. __ ..
Z
6
"0
til
o
...J
~
(j)
ro
...J
50 -
---- -- -- - ---; - -- - --- - - - - - ~ , --- -- - - - - -- --:- - - --- - -- - -- -, - - -- -- - - -- -- ,- ---- - - -- - --
---.-_.-.-.
:
-------_ .. _-,.---- - .. ---- - - - ~ + --
--Without P-O
. . . . .
With
P-O
:
:
200 180 160 140 120 100 80 60 40 20
OL.---.:.---,-----'----+----i---+---+----,-----+---l
o
Displacement (mm)
a) Lateral Load-Displacement Relationship
---- --
--Without P- 0
...... With P- 0
....... 'i ;0- ..
-- -- -- - - -- - - - -- -- - - -:- ------
-- - - -:- -- -- -------- -- -:- -- --- ----- ------ -----
500
450
- --
400
350 --
---
E
300
--- ------.+-
Z
~
'-'
.....
250 -
c
Q)
E
200
-._----- .. ---
0
~
150
100 -
50
O+----;-----;-----i----i-----i-----i-----i-----!
o 0.00005 0.0001 0.00015 0.0002 0.00025 0.0003 0.00035 0.0004
Curvature
b) Moment-Curvature Relationship
Figure-6.3.4 Moment-Curvature Analysis for Specimens-t. 2
94
- - - ---- - - - -,- - -- -- - - - - - - ~ - -~ -----
:
._----
-~ ---
~ ' - =
- -
.
" - -
.
-
. .
- -
.
-
,
- - -- ':"" .........
-
. .
.
, . ,; .
,. ......, .....: . '
0
7
:
:
I
--Without
P-O
.
-
-
-
.
With P-o
i
:
, , , , , , ,
250
200
---
Z
~
150
......,
"0
ctl
0
....I
ro
100
...
(j)
.....
ctl
....I
50
o
o 20 40 60 80 100 120 140 160 180 200
Displacement (mm)
a) Lateral Load-Displacement Relationship
:
--- --------------
:
~ ~ .
. . . . . -
. .
-- :
-
-. . . .. . .. . .
. -.
. - -
-
.. :
~
..
.
.
-
1
--Without P-O
------ ------
-
.
-
. .
-With P-o
------
: :
, , , , ,
o
o 0.00005 0.0001 0.00015 0.0002 0.00025 0.0003 0.00035 0.0004
500
450
400
350
---
E
300
z
C
.....
250
c
(j)
E
200
0
~
150
100
50
Curvature
b) Moment-Curvature Relationship
Figure-6.3.5 Moment-Curvature Analysis for Specimen-3
95
:
.. - - - . ~ - .
-------
-
.
-
.. :- .. .. -- . .
-
.... '"
~
...-.-:
.
-
.
, .
-
.
. -
~
-.
r:
/
:
--Without P-t5
. . .
--With
P-t5
:
I I I
250
200
Z
6 150
"C
ro
0
..J
iii
....
100
OJ
iii
..J
50
o
o 20 40 60 80 100 120 140 160 180 200
Displacement (mm)
a) Lateral Load-Displacement Relationship
:
---------
~ - : - : - : - -
-
.-
.
-- -
.
.'
-
.. .
.. . . . . ..
.
.
.......... ....
0
.
r
. .
I
.,
Without
P-t5
- - - * - ~
.
. .
--With
P-t5
I I o
o 0.00005 0.0001 0.00015 0.0002 0.00025 0.0003 0.00035 0.0004
Curvature
500
450
400
350
E
300
z
6
C
250
OJ
E
0
200
:2
150
100
50
b) Moment-Curvature Relationship
Figure-6.3.6 Moment-Curvature Analysis for Specimen-4
96
- - - - - - - - ---
. .
-
.. .. .. .,. o
-:-
.
-
0 . .
. .
0 .
'"
0
-
.
-
-
.
--.-
"
:
:
-----_.-------
:
--Without
P-o
-
. .
With
P-O
--------
;
I
o
o 20 40 60 80 100 120 140 160 180 200 220 240 260
Displacement (mm)
250
200
Z
6
150
-c
ell
0
....J
'"iii
l-
100 Q)
ro
....J
50
a) Lateral Load-Displacement Realationship
500
450
400
350
E
300
z
6
....
250
c
Q)
E
0
200
150
100
50
--
-
-
. .
-
. .
-
-
. . . 0
.0
-
.
-.
.
-
..
-
-
:._ ___ .. .-__- __ _ _ _ ______
.. .. :
!
---.-
-------
:
--Without P-O
-------- --
. . .
---With p-
s
o
o 0.00005 0.0001 0.00015 0.0002 0.00025 0.0003 0.00035 0.0004 0.00045 0.0005
Curvature
b) Moment-Curvature Relationship
Figure-6.3.7 Moment-Curvature Analysis for Specimen-5
97
Table-6.3.2 Theoretical Ideal Lateral Strength of Test Specimens
Ideal Lateral Strength L; (kN)
No. of specimen Moment-curvature analysis NZ3101
Without P-c5 With P-c5 Without P-c5 With P-c5
1 169.5 168.7 170.0 169.0
2 169.5 168.7 170.0 169.0
3 168.5 167.8 168.4 167.5
4 169.2 168.4 169.5 168.5
5 158.0 157.2 152.8 152.2
6.3.3 Progress of the Tests
6.3.3.1 Specimen-1
The first crack was found at the base of the column when the lateral load reached
about 60 kN. The lateral load reached 127.5 kN (75% of the ideal strength seen in Table-
6.3.2) without yielding of any main-bars, in both directions of east and west. At this stage,
twelve lateral cracks were found on each of west and east faces of the column between
50mm to 100mm apart, in the region up to about 1m above the base of the column. The
reference yield displacement 6}t was determined as 9.9mm by the procedure explained in
4.2.
When the displacement reached 46}t, the cracks had spread up to 1.5m above the base
and were spaced at every 50mm which corresponded to the location of the hoops as seen
Fig-6.3.8.
98
Figure-6.3.8 Cracks on South Face at 46),
Some diagonal cracks on the south and north faces and some vertical cracks on the
east and west faces had become signif icant when the displacement reached 6o
y
. When
the maximum displacement reached -8bj; in the west direct ion, the first buckli ng of the
longitudinal bars were detected on the west face and the cover concrete was pushed out as
seen in Fig-6.3.9(a). In the next loading cycle (2nd cycle to SOy in east), most main-bars in
the east face also buckled and the cover concrete spalled off as seen in Fig-6.3.9(b).
99 THE L1BRAR(
UNIVERSITY OF CANTERBURY
N.Z
(a) North face with buckling in west direction
(b) East face
Figures-6.3.9 Buckling of Main-bars, 8b}- (Specimen-1)
100
The location of buckling for all the main-bars was limited between the second hoop and the
thi rd hoop, 1OOmm and 150mm above the base of the column respectively.
During the cycle to 108/ in the east, the third hoop from the base of the column
fractured on the east face, at the hook of a cross-bar (Fig-6.3.10). This was caused by the
large curvature of the hoop between the hook of the cross-bar and the adjacent main-bar.
Figure-6.3.10 East face, Fracture of a Hoop at108/
The second hoop from the base of the column was broken in the west face at the
second cycle to 1Ooy to the west. At the same stage, a hook of a cross-bar was opened
and released the third hoop (150mm above the base), as seen in Fig-6.3.11. Finally, this
test was terminated after the two cycles of 120y were completed. In the second cycle to
128
y
, four main-bars were broken in tension. Fig-6.3.12 shows the condition of Specimen-1
after the test. The horizontal cracks were spread up to about 1.5m above the base and
some significant diagonal cracks diagonal seen between the horizontal cracks near to the
edges, on the north and south faces. Four hoops at 5cm to 20cm from the base were
loosened.
101
Figure-6.3.11 Fracture on a Main-bar and Opened Hook of a Cross-bar
102
(a) East Face
Figures-6.3.12 Specimen-1After Test
103
(b) South Face
Figures-6.3.12 Specimen-1 After Test
104
6.3.3.2 Specimen-2
The reference yield displacement 6:; was calculated as 9.5mm using the same
procedure as Specimen-1. A maximum displacement 12bY was applied in both east and
west directions as the second and third cycles. The maximum displacement 12bY was
chosen for Specimen-2 because Specimen-1 reached the ultimate state after the first cycle
to 125
y
.
The behaviour of the column in the first cycle for three-quarters of the ideal strength
was just the same as Specimen-1. The first crack was found at the base of the column on
the tension side (west). About ten cracks were equally distributed from the base to about
1m high.
During the second loading cycle, with the maximum displacement of 12bY, about twenty
cracks had spread up to 1.5m above the base, and a few cracks were found as well on the
base-block near to the column, at +25
y
(west). At +65
y
, some vertical and diagonal cracks
were found at the bottom part of the column on both north and south faces near to the east
face, which was in compression. At +8 bY, the widest crack at the construction joint at the
base of the column was about 5.5mm wide. When the displacement reached +12bY, some
cover concrete at the bottom on the east face had spalled off, as seen in Fig-6.3.13.
105
Figure-6.3.13 East Face (in compression) at +126j.,
When the displacement returned zero, the cover concrete buckled outwards over a
relatively large area on the west face, at 200mm to 300mm above the base, as seen in Fig-
6.3.14. It happened even though the west face was in tension because the main-bars in
the west face significantly stretched during the load cycle to 12b}t eastwards, and buckled
while the column was returning to the centre point. Some buckled main-bars were
observed, see Fig-6.3.15, when the bulged cover concrete was removed in order to avoid
damage to the potentiometers . The location of buckling on the main-bars was at from
150mm to 300mm from the base.
106
Figure-6.3.14 Bulged Cover Concrete on West Face
J
Figure-6.3.15 Buckled Main-bars (west face)
107
-.
In the latter half of the second loading cycle (0 to -125: ; in west), the 5th hoop (250mm
above the base) was broken at the hook of a cross-bar in the west face, before the
displacement reached -88y. At -125:; . many cracks were seen on the base-block, on the
tension side (East) (Fig-6.3.16) .
(a)Tension Side (East)
Figures-6.3.16 Specimen-2 at -125: ;
108
(b) Compression Side (West)
Figures-6.3.16 Specimen-2 at -125
y
On the third loading cycle, which is the second cycle to 12hY eastwards, all the main-
bars buckled at 100mm to 150mm above the base on the compression face (east). Also,
the 2nd and the 3rd hoops, at 100mm and 150mm above the base respectively, were
broken at the hooks of some cross-bars grabbing those hoops, and one hook of a cross-bar
was opened in the same face (east-compression side).
On the latter half of the third cycle to -128/ westwards, the main-bars in the
compressed face (West) buckled significantly at 100mm to 250mm above the base. In the
fourth loading cycle, namely the first cycle to 10hY, the main-bars in a side face (south)
started buckling near to the compression side (east). It led to spalling of the cover concrete
in the south face. In the latter half of the fourth loading cycle to -10hY westwards, some
cracks on a main-bar at the south-east corner were observed on the inside surface of the
bar at a buckled part, when it was extended in tension. It is assumed that this was because
the corner bars were well held by the hoops and the shorter effective buckling length
compared with inner main-bars led to large buckling curvatures of the corner main-bars.
These curvatures caused significant compression on the bar at the inside of the buckling
109
curve, and then the cracks occurred when the bar was extended in tension again.
In the second cycle to +1Oq., eastwards, cover concrete spalied off significantly also on
the side faces, north and south, near to the compression side (west). Until the test was
terminated after two cycles to 2 ~ / , the cover concrete continued spalling off gradually but
no main-bar was broken although some cracks were detected on the main-bars at the
corners.
(a) From South-West
Figures-6. 3,17 Specimen-2 After Test
110
(b) From North-East
Figures-6.3.17 Specimen-2 After Test
111
(
(d) Plastic Hinge Zone (North )
Figure-6.3.17 Specimen-2 After Test
112
In all, five hoops at 5cm to 25cm above the base were loosened. A few significant
diagonal cracks were detected in the plastic hinge zone, in the north and south faces. Fig-
6.3.17 shows the condition of column at the end of the test.
6.3.3.3 Specimen-3
The measured displacement in the north direction corresponding to three-quarters of
the calculated ultimate load, was about 50% larger than those for the other three directions.
It was found that this difference was caused by a small extra rotation of the counter-weight
basket on which the measuring frame was fixed, due to lack of high strength plaster applied
between the counter-weight frame and the base block. Therefore, the reference yielding
displacement Oy was determined as 13.3mm by averaging the other three measured
displacements, and the test was continued. The average of other three displacements,
however, was still 30% larger than that of Specimens-1, 2 and 3. The same reason can be
assumed for it.
When the specimen was loaded up to 20
y
in the east and west directions, some
diagonal cracks were found on the side faces (north and south) near to the compression
side as seen in Fig-6.3.18. These diagonal cracks were not observed on Specimen-1 with
the same displacement. It shows that additional loading in transverse directions (north and
south) caused more damage on the concrete of Specimen-3. When four cycles of 25ywere
completed, the lateral cracks were seen almost every 50mm from the joint up to about 1.5m
above the base.
113
Figure-6.3.18 Detected Diagonal Cracks (On south face, 20 to West)
Within the next four cycles to the maximum displacement of 40, a few diagonal and
vertical cracks occurred near to the base, especially at the corners of the column, as seen
in Fig-6.3.19. The width of the crack just at the base became more than 3mm at 48/.
The first buckling of the main-bars was detected during the first cycle to 60 in the east-
west direction, when the column returned from 60 west to the centre point. This buckling
occurred on the east face, and pushed the cover concrete out at 150mm above the base,
as seen in Fig-6.3.20.
114
Figure-6.3.19 Diagonal and Vertical Cracks Concentrating on an Edge
(West face, after 2cycles of 48y in E-W and one push to 48y in south)
Figure-6 .3.20 First Buckling of Main-bars (East face)
115
Buckling of the main-bars next occurred in the west face, during the second cycle to 6b}t in
the east -west direction, when the column came back from 6b}t east to the centre point. In
the following south-north loading to 6b}t, the buckling on the north face occurred in the first
cycle to 65
y
, while the column was pushed to -6b}t northwards (Fig-6.3.21). At this time, a
significant drop of lateral load was detected on the hysteresis loop. The next buckling
occurred in the south face while the column was returning from -65
y
north to the centre
point. This procedure of buckling is slightly earlier than that for the previous east-west
loading. It indicates that some damage on the column caused by east-west loading
affected the behaviour of the column during south-north loading. During the second cycle
to 65
y
, one hook of a cross-bar was opened and released the thi rd hoop, 15cm above the
base.
Figure-6.3.21 Buckling on North Face
During two loading cycles of 8b}t in the east-west direct ion, the damage at the plastic
hinge zone was advanced and the hooks of some cross-bars and hoops were opened
without breaking any hoops, as seen in Fig-6.3.22.
116
J
I
I
I
Figure-6.3.22 Damage in Plastic Hinge Zone (at 8bY)
During the first cycle to 85
y
in the north-south direction, three main-bars, one at the
south-east corner, one next to it in east face and the second one from the north-east corner
in the east face failed. This was regarded as the end of the test. One further cycle to 105
y
was applied in the east-west direction to confirm that the lateral strength of the column had
significantly decreased. Seven main-bars were broken, necked or cracked during this test.
All of them were concentrated near to the corners of the column, and between the second
and third hoops, 10cm and 15cm from the base respectively. The final situation of
Specimen-3 is shown in Fig-6.3.23. The cracking pattern of each face is similar to the
others. Cracks are spread up to 1.5m above the base, and some significant diagonal
cracks are seen near to each corner, up to about 50cm from the base. Four hoops at 5cm
to 20cm from the base were loosened.
117
(a) From East
Figures-6.3.23 Specimen-3 at the End of the Test
118
(b) From South
Figure-6.3.23 Specimen-3 at the End of the Test
119
(c) Plastic Hinge Zone (From Wast)
Figure-6.3.23 Specimen-3 at the End of the Test
120
6.3.3.4 Specimen-4
Duri ng the loading to three-quarters of the ideal strength in each direction, no
degradation was detected by the previous loading, on the stiffness of the column. Some
horizontal cracks were spread over up to about 1.2m above the base of the column. The
reference yielding displacement bY was calculated as 10.1mm.
After two cycl es of 2bY loadi ng, a few diagonal and vertical cracks were seen on all the
faces of the column up to about 50cm above the base, especially concentrating on the
corners just above the base as seen in Fig-6.3.24.
}; (,
,.
/'
-:
/
-'
I
143
5) Theoretical values of the plastic hinge zone length overestimate that for small
displacement ductility factor (say, when f..Jd=2). Otherwise, experimental values
correspond well with the theoretical values for f..Jd=::6, and particularly with Paulay and
Priestley [17]. The Japanese specification tends to overestimate L
p
(Fig-7.1.7 to Fig-
7.1.11).
6) All the values of L
p
for Specimen-5 are larger than the corresponding values for
Specimen-1. It means that the concrete strength Fe could affect L
p
(see Fig-7.1.7 and
Fig-7.1.11) even though it is not included in the equations for L
p
in both Paulay and
Priestley [17] and the Japanese specification [1].
If the plastic hinge zone length is overestimated in the design of a reinforced concrete
column, the displacement of the column for a given plastic curvature will be greater, leading
to a smaller design seismicity. Therefore, the smaller L
p
should be used for a column
designed with smaller displacement ductility factor.
7.1.4 Details of Yield Penetration
Shima et al. [24] defined a formula to obtain the extracted length of a main-bar from the
base block, Sy, due to the yield penetration, as Eq(7.1.8).
where
s, =aSy1
_ 7.48
y
(6+35008
y
)db
Sy1 - f' 2/3
c
a = 1+0.ge.45(1-aldb )
(7.1.8)
(7.1.9)
(7.1.10)
a is a factor to consider the effect of centre-to-centre distance a (mm) of the main-bars; Sy1
is the extracted length of a single main-bar (mm); 8
y
is the yield strain; db is the diameter
(mm) of the vertical main-bars; and Fe is the specified compressive strength of the concrete
(MPa). Syis assumed to be stable regardless of the strain in the main-bar after yielding.
Ogasawara et al. [25][26] actually measured Sy in a series of seismic tests with reinforced
concrete bridge columns (650
x650mm
square section) subjected to a flexure, using a steel
cable which was set in a pipe penetrating the base block and one end of the cable was
attached to a main-bar (d
b=10mm)
at the base of the column. The same loading pattern as
Fig-4.2.7 with more loading cycles than the author's test, was used. Through Ogasawara's
tests, it was observed that Sy was stable after yielding, matching well with the theoretical
144
values using Eq(7.1.8) to Eq(7.1.10) inclusive, and then stepped up when the strain of the
main-bar reached about 0.02 in the cycle of loading to 1q,. It was also observed that the
percentage of displacement (not plastic displacement) of the column resulting from yield
penetration into the column base was stable at from 10 to 15% all through their tests which
finished at about 6q,. These observations support Eqs(7.1.8) to (7.1.10) by Shima et al. at
the early stage of the tests, and also support Eq(7.1.1) by Paulay and Priestley [17]
concerning the stable influence of yield penetration on the displacement of the column.
From these findings, it is assumed that Sy is stable at a low level after the main-bars
yield, and at some stage, it increases to some higher level which is in proportion to the
displacement of the column. From the observation of Fig-7.1.7 to Fig-7.1.11, it is also
assumed that the "step point" of Sy is affected by the number of the cycles of loading. Figs-
7.1.12 and 7.1.13 show the strain history of the main-bars at 50mm below the base of the
column (50mm deep in the base block), for Specimen-1 and Specimen-4. The strain steps
up during the loading cycles to 4q, in Specimen-1 with uni-axial loading, on the other hand,
the strain steps up at the early stage of the loading cycles to 2q, in Specimen-4. It
corresponds well to the assumptions made above.
145
10000 -r---------'---------------H-----'II-""lHI"-,
7500
5000
-North East
-East
--South West
--West
-NorthWest
2500
o
=
-2500 ---- ------------ ---------- ---- ----- ------------ ---------- -- -------- -------- ---- --------- -- -- ----- -----
0.758
y
-5000 -l..- -l- 4--4!--4_4-l..-I-lI
Scan Step
Figure-7 .1.12 Strain History of Main-bars 50mm below the Base (Specimen-1)
10000 , . - - - - - - - - - - - - - - - - - - . ~ - - - ...........---..,....,..____.._.
--North East
-East
--South West
--West
Norst West
0.758
y
7500
-2500 --------------------------- ------- --- -----------------------------------
5000
c
'm
....
Ui
2500
0
....
o
~
0
-5000 ..l..- -.ll_...lI....J_---J....:....L.1.-__--l
Scan Step
Figure-7.1.13 Strain History of Main-bars 50mm below the Base (Specimen-4)
146
Sy of Specimen-1 for the early stage is calculated by the Shima's method using
Eq(7.1.8) to Eq(7.1.10), inclusive, and the numbers; db =1Omm, By =0.0015, Fe =30MPa and
a = 38mm, as
Sy1 = 7.4 x 0.0015 (6+3500
xO.0015)
10 / 30 = 0.130 mm
a = 1+0.9 e (1-38110) = 1.055
therefore, Sy= 0.130 x 1.055 = 0.137 mm
With this value of Sy and assuming the distance between the main-bars with Sy from the
neutral axis of the section at the base of the column is 440mm (from Moment-Curvature
analysis, at f.itJ. =2), the rotation of the column due to the yield penetration, Byp, is
B
yp
= 0.137 /440 = 0.000311
Therefore, the displacement of the column due to the yield penetration, b}tp, is obtained as
b}tp = B
yp
h = 0.000311 x 2250 = 0.7 mm
This value is only 7.1% of 9.9mm which is the plastic displacement of Specimen-1
corresponding to f.itJ. =2, and less than one-third of 26 % derived using Eq(7.1.1). It means,
if the behaviour of yield penetration follows Shima's equations at the early stage, actual L
p
is smaller than that shown in Fig-7.1.7 to Fig-7.1.11, which is based on the influence of
26%.
7.1.5 Proposal of Plastic Hinge Length for Japanese Specification
Only the values of L
p
corresponding to f.itJ. =2 (j.itJ. =2.74 for Specimen-3) are
recalculated with the influence of 7.1%, and shown in Fig-7.1.14 with other values of L
p
for
f.itJ. larger than 4. Each L
p
is the average of all values at the same f.itJ., for each specimen. A
revised Lp-f.itJ. relationship is proposed in Fig-7.1.14, which consists of two parts; a
horizontal line at the theoretical L
p
calculated using Eq(7.1.1) for f.itJ. more than 4, and a line
connecting a point with a quarter of the theoretical L
p
at f.itJ. =1 with the horizontal line at f.itJ.
=4. This proposed Lp-f.itJ. relationship is applicable only for rectangular reinforced concrete
147
columns which are designed using the Japanese Specification with small axial load and
assumed to fail by flexure.
350 -...----------------------------.
300
-------------------------------------0------ -. - ... -... -----.-----.------.-.----.--.--.- ..
I::.
-- . - - . . . - - - - . - --- - --- --
Specimen-1
Specimen-2
Specimen-3
Specimen-4
Specimen-5
Paulay&Priestley
Japanese Spec
'Proposed Line
CI
I::.
X
+
o
-
---------7--------------------------
_._---/---_.... _._----------_ .. __.... _-------------------
/
CI
...... -- .. -- ....... ---..---........---..---....... - .. - ...... ~ - - - - - - - X - - - ~ - - - - - . . . . _ - - - - ......._-"'"
_______________ . ._. .L . . . . .. . _ ~ _
o x/
50
100
250
E 200
.s
...,JQ. 150
10 8 2 4 6
Diplacement Ductility Factor (jJ.d )
Figure-7.1.14 Proposed Plastic Hinge Zone Length for the Japanese Specification
o+,-----+------+-----4------+-------1
o
7.2 Moment-Curvature Analysis
Table-7.2.1 shows the lateral displacement corresponding to the reference yielding
point obtained by the moment-curvature analysis and the tests. Experimental data is from
5 to 14% larger than that by the moment-curvature analysis. This range of error is
adequately small considering the size of the specimens.
148
Table-7.2.1 Reference Yielding Displacement
Reference Yielding Displacement "Y (mm)
Moment-curvature Analysis Experimental Data
Specimen-1 8.89 9.9
Specimen-2 8.89 9.5
Specimen-3 9.04 Data Error
Specimen-4 8.95 10.1
Specimen-5 9.97 11.4
Figs-7.2.1 to 7.2.5 show the hysteresis loops of lateral load-displacement relationship
for all the test specimens, with the results of moment-curvature analysis, described in
6.3.2.2.
149
Displacement Ductility Factor
-12 -10 -8 -6 -4 -2 o 2 4 6 8 10 12
150
100
Z 50
~
'-'
-100
-150
--Hysteresis
- - M-cur analysis (witI"P- 0 )
CJ CJ CJ CJ CJ 0 0 0 CJ CJ CJ CJ CJ CJ 0 0 CJ
lD -er- 0-J CJ
=
lD -e- N 0-J -er- lD
=
0 0-J -er- lD
~ ~
I I I I
I I
Displacement (mm)
Figure-7.2.1 Hysteresis loops (Specimen-1)
Displacement Ductility
-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12
200
--.....
150
100
......,
50
z
C
"'0
co
a 0
...J
"
....
co
-50
...J
-100
-150
--Hysteresis
-
- M-cur Analysis (with p: 0 )
-200
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
CD
""""
C'J 0 co CD -e- C'J C'J '<T CD co 0 C'J '<T '-0
-
~ ~
-
I I I I
I I I I
Displacement (mm)
Figure-7.2.2 Hysteresis loops (Specimen-2)
150
,
:
150
~ Z
100
E
..
50
c:
CD
0
E
Q)
(J
<ll
-50
0..
U1
i5 en
-100
-150
-150 -100 -50
W
o 50
E
100 150
Displacement (mm)
c) Loading Orbit
Figure-7.2.3 Hysteresis Loosp (Specimen-3)
151
Distlacement Ductility Factor
-8 -6 -4 -2 0 2 4 6 8
200
-
150
100
Z
6
50
"0
(\]
0
-l
0
iii
"E
0
-50 N
.
:c
-100
--Hysteresis
-150
-M-cur (withP-8
-200
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
'-0 -e- c-o 0 co '-0 -er-
""
c--, -er- co co 0
""
"<T co
-
-
-
'I
I I I I
I I I
Displacement (mm)
a) In East-West Direction
Distlacement Ductility Factor
-8 -6 -4 -2 0 2 4 6 8
200
---
--
150
100
z
6
50
"0
(\]
0
-l
0
:
c
0
-50
N
':5
:c
100
--Hysteresis
150 - M-cur Analysis (withe, 0
-200
co co co co co co co co co co co co co co co co co
co "<T C'J co co co "<T C'J C'J "<T co co co
"<T co
,.. ,.. ,.. ,.. I I I I
Displacement (mm)
b) InNorth-South Direction
--------
... --------
...,...........----......----"'--"'Y------- --------
20 -------- -------i"---:,.'---I
100 -,--------------------------,
80
60
40
'E
0
2l
(\] -20
C.
is -40
-60
80
100 50 o -50
100 +------+-----+------+-------1
-100
Displacement (W-E)
c) Loading Orbit
Figure-7.2,4 Hysteresis Loops (Specimen-4)
Hi?
Displacement Ductility Factor
-10 -8 -6 -4 -2 0 2 4 6 8 10
200
-
- -
150
100
Z
6
50
"C
ro
0
....J
0
(ij
-
c
0
. ~
-50
0
:c
-100
--Hysteresis
-150
-
-M-cur Analysis (Withp-a )
- - -
-200
CJ CJ CJ CJ CJ CJ CJ CJ CJ CJ CJ CJ CJ CJ CJ CJ CJ
co "'7 C'J CJ CD co "'7 C'J C'J "'7 to CD
=
C'J "'7 <.D
~ ~
- -
I I I I
I I I I
Displacement (mm)
Figure-7.2.5 Hysteresis loops (Specimen-5)
The moment-curvature analysis estimates the columns to sustain their lateral strength
at a higher level than the experimental results for displacements more than 180mm (except
for Specimen-5). However, the test results of Specimens-1, 3, 4,and 5 show some
degradation of strength from cyclic loading. Cyclic loading resulted in buckling of the main-
bars and the failure of the transverse reinforcement, it caused loss of confinement, and
then the lateral strength of the column decreased. This procedure of column failure was
accelerated by the additional loading cycles to the same displacement in Specimens-3 and
4, as seen in Figs-7.2.3 and 7.2.4. It can be seen in Figs-7.2.3 and 7.2.4 that the lateral
strength of the column in the north-south direction is always smaller than that in the east-
west direction with the same displacement, because the displacement in the north-south
direction was always applied after the east-west direction. Specimen-2 sustained its lateral
strength without degradation, at approximately the same level as the result of the moment-
curvature analysis in the first cycle to 120:" eastwards. The main-bars did not buckle and no
transverse reinforcement failed in this loading cycle to the maximum displacement.
The moment-curvature analysis shows that the lateral strength has a plateau after
yielding and strain hardening leads to a 15% increase in strength after the displacement
153
reaches about 40mm. In the test results, however, the lateral strength increases after
yielding to the level of the strain hardening, suggesting that the strain hardening occurs
earlier than the moment-curvature analysis. Figs-7.2.6 and 7.2.7 show the strain history of
the main bars of Specimens-1 and 2, at 50mm above base of the column. It can be seen
that the strain of some of the main-bars goes up to the measurement limit (about 0.02)
during the loading cycle to 2b}t in both Specimens-1 and 2.
20000
17500
15000
12500
10000
c
'(0
7500 ...
Ul
e
5000 o
2500
0
-2500
-5000
-7500
North East
-East
--South East
--South West
"
North West
co
C',
..__ . -_...---__ --D..-7S3
y
- -- ---__ --__ - __ - . ---2 -- _. - -- -A'O-y _
Scan Step
Figure-7.2.6 Strain of Main-bars at 50mm above the Base (Specimen-1)
154
25000 -...-----------------------j-------,
20000
. ------- - -- - - - - - _ ~ ~ _.---- ---- - - - ~ _ - - -- -._---- ---- _.. _ - _ ~ _ - - - - - ---- -- - --... - _. ------_ .. ~ - - - - - - - ---------- -------- -----.
15000
c
'(0 10000
...
ii5
e
~ 5000
-SouthWest
-West
-NorthWest
-5000
----------- -- ---- -- -- -- -- ---- -- 0.758y --------- ---- ---- -------------- ---- ----------- ---25;,---- ------ -------
-10000..L-------------------------...J
Scan
Figure-7.2.7 Strain of Main-bars at 50mm above the Base (Specimen-2)
This behaviour is assumed to be due to the higher strain on the main-bars concentrating at
the cracks. For a given plastic rotation or a displacement, the strain of the main-bars and
the plastic curvature of the section used in the moment-curvature analysis are defined as
the average of the value varying through a plastic hinge length, for convenience. At a
crack, however, the strain of the main-bars must be higher than that of the same main-bars
in concrete. Therefore, the tensile stress of the section due to flexure concentrates on the
main-bars once the section cracks, and the strain hardening could occur much earlier in the
cracks than estimated by the moment-curvature analysis. Based on this assumption and
the plastic hinge zone length proposed in 7.1, the moment curvature analysis was carried
out with early onset of strain hardening and the revised relationship of lateral load-
displacement and moment-curvature are shown in Figs-7.2.8 and 7.2.9. The revised line
matches well to the experimental result of Specimen-2. In the calculation of the revised
moment-curvature analysis; 0.002 was used for the strain of the main-bars corresponding
to onset of the strain, which is about one-fifteenth of the measured value seen in Fig-6.2.2,
because the strain of main-bars usually increases suddenly to the measurement limit once
the bar yields, as seen in Fig-7.2.6. Since the proposed plastic hinge zone length in Fig-
7.1.14 varies depending on the displacement, the revised moment-curvature analysis was
155
carried out by repeating the calculation equilibrating the plastic hinge zone length and the
displacement. for JLL1 less than 4.
156
250 - r - - - - - - - - - - - - - - - - - - - - - - - - ~ - - - - _ _ ,
200
--
. .-
. .
..
.; ~ . - - . - - - ~ ...,-.-.... ---- ....... - .-- ... ------..-,.-,,- ...-,-.-.. -..;. ; - ~ ~
Z
6 150
u
ell
o
....J
~
s 100
ell
....J
--Without p- 0
..... With P-O
50
200 180 160 140 120 100 80 60 40 20
0-!----+----+------:---+----i----+----+---+---+-----1
o
Displacement (mm)
a) Lateral Load-Displacement Relationship
:
:
--
.i->:
.
.. . .., .
.
. , .... _...., .
.... .... :.......... :
./'
. ..
-----
--- ----- -- -----+
--_.----
:
--- --
-----_.- -- --
------
--Without
p-
s
:
. . . . . .
With p-
s
----- -- ----- - ~ - - +- ---- -- ----- +-++--
---
;
500
450
400
350
E
300
z
6
C
250
Q)
E
0
200
~
150
100
50
o
o
0.00005 0.0001 0.00015 0.0002 0.00025 0.0003 0.00035 0.0004
Curvature
Moment-Curvature Specimen-1,2
Figures-7.2.8 Result of Revised Moment-curvature Analysis for Specimens-1 and 2
157
Displacement Ductility Factor
-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12
250
200
150
100
Z
50 ~
.......-
"'0
(I]
0 0
-I
(ij
...
-50
ro
-I
-100
-150
-200
-250
co co co co co co co
(0 -er- 0-J co
=
(0
-st-
~ ~ ~
I I I
I I I
-----
--Hysteresis
- M-cur Analysis (with p: 0 )
co co co co co co co co co co
~ ~ 7 ~ ro 0 ~ ~ ~
I
Displacement (mm)
Figure-7.2.9 Revised Lateral load-Displacement Relationship
with Hysteresis loop of Specimen-2
7.3 Energy Dissipation
7.3.1 Calculation of Dissipated Energy
The energy dissipated by a test specimen is defined as the area enclosed by the
hysteresis loop of the lateral load-displacement relationship, with units of kNmm. The
calculation is carried out by a spreadsheet, numerically adding up the area in a loop for
each completed loading cycle, up to the cycle before the specimen reached the ultimate
state, described in 5.2.2.1. For Specimen-3, the measured displacement was adjusted by
subtracting the difference between the measured yield displacement of Specimen-3 and the
average yield displacement of Specimens-1, 2 and 4, in order to eliminate the effect of the
extra rotation of the measuring frame, explained in 6.3.3.3.
158
7.3.2 Individual Cycle Energy
Figs-7.3.1 to 7.3.5 show the dissipated energy by each individual loading cycle and the
accumulation of the dissipated energy. In Fig-7.3.3 for Specimen-3, one column in the
graph is the energy dissipated by two loading cycles to a given displacement in the same
direction. In Fig-7.3.4 for Specimen-4, one column is for half of a completed double-8
shaped loading cycle (one 8 shape), seen in Fig-5.2.5(c).
35000 250000
c=:J Dissipated
Q)
30000
Energy
hoop failure
13
200000 6; >.
o
......... Cumulative
....
Q)
.c
25000
c:
o
Dissipated
w
ell
w
Energy
"0
....
150000 "* _
0
........
20000
-
E
ate >. E
z
Q)
15000
c:-
100000 (\) -
W
"'0
>
:;:;
Q)
10000
ell
.....
"3
ell
.9- E
(J)
50000
::J
(J)
o
is 5000
c-, c-, c-,
.>. c-, .>, c-, o-, -, .>, c-, c-, c-,
-0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0
en
'"
'"
"'T "'T '-0 '-0
=
co co co
'" '"
r----
------
<;
------ ------
co
'"
c--r c--r c--,
------
------ '" '"
Loading Cycle
Figure-7.3.1 Individual Cycle Energy (Specimen-1)
159
D
i
s
s
i
p
a
t
e
d
E
n
e
r
g
y
f
o
r
E
a
c
h
C
y
c
l
e
D
i
s
s
i
p
a
t
e
d
E
n
e
r
g
y
f
o
r
E
a
c
h
C
y
c
l
e
(
k
N
m
m
)
(
k
N
m
m
)
-
>
.
I
\
.
)
W
.J
:>
..
0
1
0
'>
-
>
.
I
\
.
)
W
.J
:>
..
0
1
0
'>
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
7
5
d
y
(
E
-
W
)
~
I
I
I
I
I
I
0
7
5
d
y
I
I
I
I
+
0
I
I
1
2
d
y
t
o
"
0
7
5
d
y
(
N
-
S
)
t
o
"
C
m
o
o
m
o
C
2
/
l
2
d
y
:
:
;
3
.....
.....
o
r
o
(
I
)
~
w
"
C
:
:
:
J
-
-
(
I
)
I
C
D
~
I
o
-
"
-
o
J
.....
~
"
3
-
o
J
-
0
:
:
:
J
W
2
d
y
(
E
-
W
)
(
O
-
o
c
~
-
e
"
w
1
0
d
y
,
'
<
r
o
-
'
<
r
o
o
r
O
'
"
w
.
.
.
.
0
)
.
.
.
.
N
3
Q
-
"
0
)
(
I
)
!
:
:
!
:
C
D
c
.....
:
:
:
J
2
d
y
(
N
-
S
)
0
.
<
0
.
:
:
:
J
2
/
1
0
d
y
0
)
r
o
@
g
-
o
,
C
D
0
.
-
"
0
:
:
:
J
'
"
<
"
<
-
r
o
-
C
I
l
0 '
"
a
:
r
a
:
0
8
d
y
0
)
0
S
"
C
0
C
0
)
C
il
:
:
:
J
~
~
4
d
y
(
E
-
W
)
~
0
.
(
0
-
>
.
0
S
-
o
S
"
2
/
8
d
y
0
'>
'
<
(
0
'
<
(
0
0
0
0
0
(
I
)
0
4
d
y
(
N
-
S
)
(
I
)
'
<
6
d
y
m
~
0
r
n
C
D
:
:
:
J
C
D
:
:
:
J
(
I
)
6
d
y
(
E
-
w
)
C
D
2
/
6
d
y
.....
.....
+
(
0
3
(
0
'
<
'
<
.
.
.
-
O
l
e
n
5
"
C
i
)
4
d
y
-
0
6
d
y
(
N
-
S
)
,
-
0
m
o
o
r
n
o
(
I
)
0
"
(
I
)
~
w
"
c
:
:
:
J
-
"
O
l
C
D
~
0
.....
0
2
/
4
d
y
.....
~
_
3
3
"
0
"
3
"
.
.
.
,
-
-
(
O
-
o
c
(
0
-
0
(
I
)
8
d
y
(
E
-
w
)
C
(
I
)
'
<
r
o
-
'
<
r
o
:
:
:
J
o
:
:
:
J
.
.
.
.
r
o
.
.
.
.
2
S
:
2
d
y
C
D
!
:
:
!
:
C
D
,
:
:
:
:
r
,
,
0
.
<
W
5
"
I
\
.
)
0
.
'
-
'
"
0
'
-
'
"
C
D
8
d
y
(
N
-
S
)
0
"
(
Q
.
.
2
/
2
d
y
N
:
1
-
-
>
'
C
'
0
0
0
1
-
>
.
I
.....
I
\
.
)
I
\
.
)
N
.J
:>
..
0
'
>
,
c
o
-
>
.
-
>
.
-
>
.
-
>
.
0
(
1
)
0
0
1
(
1
)
0
0
1
0
0
0
0
0
I
\
.
)
.J
:>
..
0
'>
0
~
o
0
0
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
0
0
0
0
0
r
o
g
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
C
u
m
u
l
a
t
i
v
e
D
i
s
s
i
p
a
t
e
d
E
n
e
r
g
y
C
u
m
u
l
a
t
i
v
e
D
i
s
s
i
p
a
t
e
d
E
n
e
r
g
y
(
k
N
m
m
)
(
k
N
m
m
)
250000
re
200000
>.
e>
ultim te state ~
150000 :$
co ........
,90 E
:z E
o ~
100000 g; "'-'
:.0::-
ro
"S
E
:::l
50000 0
0
c-, .>, c-, c-, c-, c>, o-, '>,
-0 -0 -0 -0 -0 -0 -0 -0
'" '"
" " LO LO 0Cl 0Cl
-----
----- ----- -----
c--,
'"
c--,
'"
Loading Cycle
c::::J Dissipated
Energy
-+-Cumulative
Dissipated
Energy
--..--------------main-bar buckling 50000
45000
(])
40000
U
>.
0
35000
.c
o
co
W
30000
...
0
........
-
E
>. E
25000
e:'z
(]) ~
c"'-'
w
20000
"0
(])
.....
15000
co
c.
'00
(/)
0
10000
5000
0
-,
-0
en
r-:-:
0
Figure-7.3.4 Individual Cycle Energy (Specimen-4)
30000 -r---------------------....,- 180000
160000
140000 >.
e>
(])
120000 tfi
ate s t a t ~
.....
100000 [ E'
'00 E
, ~ z
80000 0 ~
g;"'-'
~
"S
E
:::l
o
....
hoop failure
if
main-bar buckling
I
~ ~ ~ ~ ~ ~ ~ ~ ~
~ ~ ~ ~ ~ u v v u
~ ~ ~ ~ ~ ~ ~ ro a
----- ----- ----- -----
~ ~ ~ ~
Loading Cycle
Figure-7.3.5 Individual Cycle Energy (Specimen-5)
c::::J Dissipated
Energy
-+-Cumulative
Dissipated
Energy
5000
15000
10000
20000
25000
(])
U
>.
o
.c
o
co
W
...
.EE'
>. E
e:'z
(]) ~
c "'-'
W
"0
(])
.....
co
c.
'00
(/)
o
161
The second loading cycle to the same displacement always dissipates less energy than the
first cycle due to the damage caused by the former cycle, as seen in all Figs-7.3.1 to 7.3.5.
A significant drop of energy dissipation is observed after the buckling of main-bars and the
failure of transverse reinforcement. The allowable ductility factor of the prototype column
for the Type-2 earthquake is 6 as calculated in 5.3.4.5. It is good news that all of
Specimens-1, 3 and 4 (even 5) sustained a good energy dissipation until at least the
loading cycles to 661 were completed, even though the bi-directional loading for
Specimens-3 and 4 caused early buckling of the main-bars and the failure of transverse
reinforcement. Buckling of the main-bars and failure of transverse reinforcement in
Specimen-2 occurred in the first cycle to 1261 eastwards and it caused a significant drop of
energy dissipation from the following loading cycle, as seen in Fig-7.3.2. This drop is also
observed in Fig-7.2.2. For the loading cycles to lower displacements, the loading loops are
pinched due to the low lateral strength and do not dissipate as much energy as that for the
other test specimens with the same displacement.
7.3.3 Total Dissipated Energy
Fig-7.3.6 shows the comparison of the total dissipated energy until the ultimate state is
reached, for all the test specimens.
162
120%
100.0%
100%
----------- ..
........
::R
78.3%
0
'-"
80% .....
I
C
Q)
E
'(3
60%
Q)
0-
CJ)
.....
(fJ
c
40%
'm
0)
20%
0% '
103.4%
102.5%
-7t:6%
Specimen-1 Specimen-2 Specimen-3 Specimen-4 Specimen-5
Figure-7.3.6 Total Dissipated Energy
The total dissipated energy of Specimens-1, 3 and 4 are approximately the same, meaning
that the bi-axial load did not affect the total energy dissipation capacity of a column until the
ultimate state. This finding would agree with the conclusion stated by Ohno and Nishioka
[27] that the total dissipated energy by a column is independent of the loading sequence.
Specimen-2 dissipated much less energy than Specimens-1, 3 and 4, but the ultimate state
defined by Zahn et al. [18] can not be applied in this case because the loading pattern was
opposite to that for Specimen-1.
When the first cycle to 10l5Y was completed on Specimen-2, however, some cracks
were found on the main-bars showing a significant damage on the column. Therefore, no
more significant energy dissipation capacity is expected, and the total dissipated energy
shown in Fig-7.3.6 could be recognised as the capacity of Specimen-2 when the opposite
loading pattern of that for Specimen-1 is applied. This means that the loading pattern
applied to Specimen-2 affected the total energy dissipation capacity of the column, causing
the buckling of main-bars and the failure of transverse reinforcement, and decreasing the
shear strength at the base of the column, at an early stage of the test.
163
The dissipated energy by Specimen-5 is the lowest among all the five test specimens.
Even though the ideal lateral strength of the specimen was estimated as 90% of that of
Specimen-1 as seen in Table-6.3.2, energy dissipation capacity was only 71.6% of
Specimen-1. It shows that the effect of concrete strength on the energy dissipation
capacity of the column is more significant than its effect on the ideal strength of the column.
7.4 Damage Index
7.4.1 Calculation of Cumulative Damage
The cumulative damage of the test specimen during the test, D, was analysed using
the fatigue model explained in 4.3 using the constants suggested by both Mander et al.
and Kunnath et al. shown in Eq(4.3.2) and Eq(4.3.8), respectively. Figs-7.4.1 to 7.4.7 show
the transition of the damage index 0 during the tests. All the graphs are plotted until the
ultimate state defined by Zahn et al. [18] except Specimen-2. As explained in 4.3, the
cumulative damage 0 (Damage Index) takes a value from 0 to 1.0. The member is
assumed to have no damage at 0=0, and 0=1.0 corresponds to the failure of the member.
The inclination of the graph becomes much steeper as the test proceeds, except for
Specimen-2 when 0 reaches 1.0 (failure) earlier, at the third loading cycle (,u.1=10) for
Mander's model and at the fifth loading cycle (,uL1=8) for Kunnath's model. The first two
loading cycles to 0.75OY are not counted in the number of loading cycles because the
fatigue model assumes that the strain of a reinforcing bar in its elastic range does not
cause any damage to the bar. This means that the loading cycles with large displacements
(large strain amplitude for main-bars) govern the damage index 0 if the number of the
loading cycle is the same for each displacement.
164
12
I
I
,
10 11 9 8 7 6 5 4 3 2 1
: :
, , : ' : )( ,I
1/128,
, ' , , , , y , ;' ,
-. ------ ----------f ----- ---- i- -- ---- -- ---- --- -i-------- +----7:-- ---- ----V---- --! ---- ---- -+-- --- --------- -- --
: : : : : :......: "', : :
: : : : :..)r : :
....... .
o I I I I I I I I I I
o
0.2
0.4
<l>
0)
0.8
co
o
0.6
1.8
:
1.6
- Mander's model
--
1.4
....-
Kunnath's model
1.2
- -
X
<l>
"'0
C
1
No. of Loading Cycles
Figure-7.4.1 Progressive Damage of Specimen-1
2-..-------:---:-----:---:----------:--------:---..
13 12 11 10 9 8 7 6 5 4 3 2 1
1.4
, : /.... , ,
;TV,r"c;....
0.6 : -Mander' --...;---------
0.4 - ----JC- ------- - .---------.----------:----------- - Kunnath's model ---- -.---------
0.2 f- -lZ------.:- ---- -- -- -j- --- ---- -- ---- ---;------ --- -:-- ------- -- -- ----- i - - - - -- - - - -i- ---------:- --------- -.------+ --------
o I.f'
o
:::nnnjnnrnntnnrn;CJ':4:tJrJn
: : : /: : : : : : :
---- - 1/1OOy:'---------;---- .---- -- ---!--------- -,-- ---- ---,----- -----:----------j------- ---:----- ----"---------
..nnmn.If: .nnn]. "mm.
, V: : "': : : : : ' ,
1.2
"C
c
OJ
OJ
co
E
8 0.8
No. of Loading Cycles
Figure-7.4.2 Progressive Damage of Specimen-2
165
15 14
I
/
13 12 11 10 9 8 7 6 5 4 3
Kunnath's model
....Mander's model
2
,
a
1.6
1.4
1.2
x
1
<Il
"0
.E:
<Il
0.8
0'1
(lJ
E
(lJ
Cl
0.6
0.4
0.2
No. of Loading Cycles
Figure-7.4.3 Progressive Damage of Specimen-3 (Comer-Bar)
1.2 -,..------;---;---,--;----,---,--,-----;-------,
I
9 10 11 12 13 14 15 8 7 6 5 4 3 2 1
I
, " "I!
.=
, , , , , , , , , , , , 'I
-------. --------:- ---- ---:- -- --- --i- -------j-- ---- -- ------ f-------+- ----- -:- -------+- --- ---i- ---- --- --1---i---;- -- ----- ---
::::::::: L-l-Jji
: : ' : : : : : : I : l
: : : / : -- : :
, , , : , : : :" , : : : :
---------
: : : : -;- : "j : : j : :
, , , / .. V, , , , , ,
..t""""'; I ,; i:::
0+--...... .......-""'\""=--+--+--+-__+--+-_+_-_+_-+_-+_--,f_____1
a
0.2
0.8
x
<Il
"0
.E:
<Il
0;6
0'1
(lJ
E
(lJ
c
0.4
No. of Loading Cycles
Figure-7.4.4 Progressive Damage of Specimen-3 (Centre-Bar)
166
1.2
1-1----'---------------+----'-----;.-----'---+----+---..;---1
0.8
x
Q)
"'0
c
Q)
0.6
OJ
ro
E
ro
0
0.4
0.2
, ,
- -- - - - -- - i- -- - -- - - --- --- - --- - - - -;--- -- ---- - - - - - - - --- - --;- -- - - - --- - - i - -- - - - ---- -f - - - -- - - - - -,-- -, --. - --, - --. -", - - -;-. - - --- - ---I
, '
- Mander's model
1 2 3 4 5 6 7 8 9 10 11
No. of Loading Cycles
Figure-7A.5 Progressive Damage of Specimen-4 (Comer-Bar)
1.2 -,----------------------------.....,
8 7 6 5 4 3 2 1
- Kunnath's model
- Mander's model
---- --- -- .-.- -- -- .---- -----. .--- --------------f---------------f---- --- --------
"
: )/
-- ---- ;----- -- _.c __ -- -- -,--- ;. -L-i----------
--""': --;"....-'
, ,,-- '
, - .. --,
_ - ---
O-l---"""'i---"""'"'"!'--...li!.....-=::;:=:.---l----+---+----l----l
o
0.2
0.8
x
II)
"'0
C
II)
0.6
OJ
ro
E
ro
0
0.4
No. of Loading Cycles
Figure-7.4.6 Progressive Damage of Specimen-4 (Centre-Bar)
167
1.2 -.---;------;---:----;--;-------; -:-- ---.
/
10 9 8 7 6 5 4 3 2
- Mander's model
.. Kunnath's model
I
/
----- _
. . . /!
, , , I
: : : / : /
................................ .T7'; .
- L : ----.------------- __ ..-- .__ ._.
. : I y
: : : : './' /: :
- - - - - - - -- -.- - -- - - - -.-. - c- - - - - - - __ ;;;./.__ __ y __ .__. .. --- .. _
. / : : :
--: .. ' . : j
.:- =-.-:-- .. ::
,
.........
a
0.2
0.8
x
Q)
""C
.=
Q)
0.6
en
ro
E
ro
0
004
No. of Loading Cycles
Figure-7A.7 Transition of Damage Index (Specimen-5)
In the calculation of 0 for Specimens-i, 2 and 5, the plastic displacement t5
p
was
obtained by just the same way as used in 7.1. The member length h=2250mm, the plastic
hinge zone length L
p=247mm
calculated using Eq(7.1.1) suggested by Paulay and Priestley
[17] and the centre-to-centre distance of the main-bars, d =494mm, were used.
For Specimens-3 and 4, the cumulative damage of the main-bars during the test
depends on the location of the main-bar. Therefore, both a bar at a corner of the section
and a bar at the centre of a face of the column were examined.
For Specimen-3, the number of the strain cycles with the same amplitude is four for a
corner-bar as shown in Fig-7A.a. A centre-bar, however, has two strain cycles with the
same amplitude as that for the corner-bar, and four additional cycles with lower strain
amplitudes (see Fig-7A.9). These additional four strain cycles are applied to the centre-bar
due to the distance between the bar and the neutral axis of the section, during the two
loading cycles in the direction parallel to the face of the column with the centre-bar. d=
200mm was assumed for the calculation of the additional low strain amplitude cycles,
because the distance between the centre-bar to the neutral axis of the base section was
168
about 200mm according to the moment-curvature analysis. Assuming the additional strain
is only in tension, the plastic strain, which is half of the strain amplitude, can be obtained by
Eq(4.3.4) with d =200mm.
169
c
'f2
.....
(f)
Scan
a) Actual Strain History
Scan
b) Idealised Strain History
Figure-7A.8 Strain History of Corner-bars in Four Loading
Cycles
170
.5:
~ ------- -------------- ----
U5
Scan
a) Actural Strain History
c
'i
U5
Scan
b) Idealised Strain History
Figre-7.4.9 Strain History of Centre-bars in Four Loading
Cycles to the Same Displacement (Specimen-3)
171
The strain history of the comer-bar and the centre-bar of Specimen-4 in one complete
loading cycle (see Fig-5.2.5(c)) are shown in Figs-7.4.10 and 7.4.11, respectively. As can
be seen, the comer-bar has one strain cycle with larger amplitude and the other two strain
cycles with smaller amplitude, and the centre-bar has only two strain cycles with the same
amplitude in one completed loading cycle.
172
Scan
a) Actual Strain History
c
. ~
iJ)
Scan
b) Idealised Strain History
Figure-7 .4.10 Strain History of Corner-bars in One
Completed loading Cycle (Specimen-4)
173
c
. ~ ------ -- -----
en
Scan
a) Actual Strain History
c
. ~
.....
lJ)
b) Idealised Strain History
Figure-7.4.11 Strain History of Centre-bars in One
Completed Loading Cycle (Specimen-4)
174
N, of the first strain cycle of the corner-bar with larger amplitude was calculated using b
p
, L
p
and d measured in the north-east and south-west direction. b
p
and dare obtained as
d(45)=.fi x 494 = 699mm
where, b
px
and b
py
are the plastic displacement for x and y (east-west and north-south)
direction, respectively. L
p
for 45 degrees was obtained as seen in Fig-7.4.12 using the
experimental data. From Fig-7.4.12, L
p
for 45 degrees can be assumed to still follow
Eq(7.1.1) by Paulay and Priestley [17]. Therefore, L
p
=247mm was used also for the 45
degree direction. The smaller strain amplitude of the corner-bar and the strain amplitude of
the centre-bar were assumed to be the same in the same loading cycle, and calculated
using the same values of L
p
and d used for Specimen-1, 2 and 5.
8
6 4
Displacement Ductility Factor (p J)
2
--- --- -- --
-- -- -- --
--
-- -- -- --- ------
-- -- --- -- --- ~ --- -- -- -- --
--
-- ----
- .
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
.
-
JI-=-.....-
+
=-.....-
;",;",;; c,.;,:;.,,- -..;;.,;;,
-
-c,;..; -..;;.,;;" ';.;;.,;
.;;..;;.,.- -..;;.,;;"
- -
-..;;.,;;" --...-
- - - - -
--...
--
.-
- - -
-.---..;;.,;;" ....... ;;..;..--..;;.,;;"
--
-- -- -- -- -- ------ ---+- -- --------- ---- ---
-- ------
II
-- -- -- -- -- ------ -- --- -- ---- --- --- -- --- -- ---- -- --
--
+
+ South East
-- --
-- -- ----- -- --
------ ------
II North West
- - - Paulay and Priestley
-- --
-- --- --- -- -- ---
--
-- -- ---------- -- --
-
.
-
.
Japanese Specification
: o
o
50
100
400
150
250
300
350
E
.s 200
S
Figure-7.4.12 Plastic Hinge Zone Length for 45 Degree Axis (Specimen-4)
175
7.4.2 Applicability of Damage Model
The findings from Figs-7.4.1 to 7.4.7 are follows.
1) The Mander's model overestimates the damage for all the test specimens except
Specimen-4.
2) The Kunnath's model estimates the failure of Specimens-1 significantly well.
3) For Specimen-2, both of the fatigue models predicted the failure at an early stage,
although none of the main-bar failed all through the test. However, some cracks were
found on a corner-bar during the second cycle to 108y, which was the 4
th
loading cycle in
Fig-7.4.2. Because the definition of the main-bar-fracture is the crack initiation for
Mander's model, Mander's model predicted well the main-bar-fracture of Specimen-2, in
this sense.
4) The damage of the corner-bar is more severe than that of the centre-bar for Specimen-3
according to both of the damage indices by the Mander's model and the Kunnath's
model. The damage index of the corner-bar calculated using the Kunnath's model
corresponds well to the actual failure of Specimen-3, and the Mander's model agrees
with the centre-bars.
5) The damage of the corner-bar is critical also for Specimen-4, but both of the damage
models underestimate the damage of both the corner-bar and the centre-bar.
6) The Mander's model predicts the failure of Specimen-5 well, on the other hand, the
Kunnath's model underestimates the damage of Specimen-5 even though it predicts the
failure of Specimen-1 well.
Kunnath's model predicts well the failure of af1exurally reinforced concrete column with
a low axial load and seismically detailed using the Japanese specification as well as that by
CALTRANS or AASHTO, if the standard loading pattern (Park [2]) in one direction is used
in the test. It also predicts well the failure of a column loaded independently in two
directions perpendicular to each other, calculating the damage index for the corner-bars. In
some cases, however, it did not predict the failure of the column well.
The mispredictions for the failure of a reinforced concrete column by the Kunnath's
model could be explained by reason of the difference in the constants of Eq(4.3.2) and
Eq(4.3.8). Mander's model was formulated as Eq(4.3.2) from experiments on steel bars to
which some cyclical axial loading were directly applied. On the other hand, Eq(4.3.8) of
176
Kunnath [6] was derived from experiments using some reinforced concrete columns.
Therefore, Kunnath's equation accounts for some effects as a composite member under a
given condition, for instance the accumulated damage due to shear, axial stress and loss of
confinement. It implies that the underestimation of the damage on Specimens-4 and 5 by
Kunnath's model is due to the earlier confinement failure compared to the damage on the
critical main-bars of the test specimens. The first spalling of cover concrete was observed
during the loading cycle with the maximum displacement of 6b}- in both Specimens-3 and 4
due to the main-bar-buckling, and the confinement failure followed. The accumulated
damage on the comer-bars, however, were different for Specimens-3 and 4 at the same
stage due to different loading patterns. It means that the difference of the loading patterns
caused the underestimation of the failure for Specimen-4. And it can be assumed that the
weak cover concrete of Specimen-5 led to the earlier main-bar-buckling and confinement
failure compared to Specimen-1, which was loaded in the same way with Specimen-5.
The findings and assumptions above suggest that the loading pattern and the concrete
strength could affect the prediction of the failure of a reinforced concrete column by the
damage models based on the fatigue of the vertical main-bars.
7.5 Proposal of Failure Prediction Procedure
The analysis in 7.3 and 7.4 imply that the failure of a column results from either the
main-bar failure or confinement failure, and the balance of them is affected by the loading
pattern applied to the specimen and the concrete strength. The fatigue based damage
model suggested by Kunnath et al. [6] predicted well the ultimate state of columns
(Specimens-1, 2, and 3) when the fatigue of the critical main-bars (corner-bars) is relatively
greater when compared with the dissipated energy. On the other hand, the failure of a
column (Specimen-4) corresponds well to the cumulative dissipated energy equal to that of
the same column cyclically loaded by the standard loading pattern suggested by Park [2],
when the fatigue of the critical main-bars is relatively low. From these findings, a procedure
to predict the failure of a reinforced concrete column subjected to an arbitrary cyclic loading
pattern is suggested as follows.
1) The lateral strength of the column and the curvature at the base of the column
corresponding to the reference yield point are calculated using the moment-curvature
analysis suggested by Mander et al. [23].
177
2) The elastic displacement of the column corresponding to the reference yield point is
calculated by the lateral strength and curvature obtained in (1) and determined as 16Y.
The maximum displacement applied to the column for each load cycle of the standard
cyclic loading pattern suggested by Park [2], is chosen.
3) The failure point, corresponding to the ultimate state defined by Zahn et al. [18] of the
column in the standard cyclic loading pattern, is predicted by the damage index
calculated by the Kunnath's damage model.
4) The total dissipated energy until the failure point is obtained using some existing
hysteresis loop model (e.g. Kunnath and Reinhorn [28], Otani [29] etc.)
5) For an arbitrary cyclic loading pattern, the cumulative dissipated energy is calculated by
the hysteresis loop model used in (4), and the damage index is calculated by the
Kunnath's model. If a bi-directionalloading is used, the cumulative energy is the sum of
each cumulative energy independently calculated in the transverse and longitudinal
directions, and the damage index is calculated for a critical main-bar (corner-bar) of
which the total strain amplitude in a load cycle is the largest in the column.
6) The failure point of the column in the arbitrary cyclic loading pattern is decided as the
earliest point where the cumulative dissipated energy reaches the total dissipated
energy in the standard loading pattern (calculated in (4)) or the damage index reaches
1.0.
The prediction of failure of Specimens-1 to 5 inclusive are carried out following the
procedure (1) to (6) inclusive described above, and Table-7.5.1 shows the predicted
maximum displacement ductility factor before the ultimate state is reached. The analysed
results for Specimens-1, 3, 4 and 5 correspond well to the experimental results. Figs-7.5.1
to 7.5.10 inclusive show the analytical results for damage index and energy dissipation of
the test specimens.
178
Table-7.5.1 Last Load cycle before Ultimate State
Last Load cycle before Ultimate State
Proposed Analysis
Damage Index Dissipated Result Experiment
(Kunnath's) Energy (smaller)
Specimen-1 (standard) 2/12oy 2/12oy 2/12oy 1/12oy
Specimen-2
Close to failure Doesn't fail
2/12oy
Specimen-3
2188y 1/88y 1/88y 1/88y
Specimen-4
1/108y 2180y 2188y 2188y
Specimen-5 (standard) 2/108y 21108y 21100y 1/10oy
Example: 2/8oy=the second cycle to 80y
179
:
1.4 +--------:-,
1.2 ------
...... - Mander's model
...... Kunnath's amodel
)
r
-'
I 21120
y
X
<Il
"0
1
E
<Il
Cl
(II
0,8
E
(II
0
0.6
0.4
0.2
0
0 1 2 3 4 5 6 7 8 9
I t
10 11 12 13
No. of Loading Cycle
Figure-7.5.1 Damage Index of Specimen-1 (from Moment-Curvature Analysis)
200000 ~
....
Q)
c
ill
'0
150000 2
ttl----
.9- E
lZ E
( 5 ~
100000 ~ -
~
ttl
"'5
E
50000 <3
,..-------------------.....,-250000
0
0
c-, c-, c-, c-, c-, c-,
""
c-, c-, c-, .>, c-,
-0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0 -0
01 01 -cr- -q- '-0 '-0 OCJ OCJ
S S ~
01
-------
------- ------- -------
01 01 GJ 01
------- -------
01 01
Loading Cycle
35000
30000
c:::J Dissipated
Q)
Energy
(3
>.
o
.s:: 25000
........ Cumulative
o
ttl
Dissipated
ill
....
20000
Energy
0----
'+- E
>. E
E'z
Q) ~
15000 1::-
ill
"0
Q)
....
10000 ttl
a.
'iii
II)
(5
5000
Figure-7.5.2 Energy Dissipation of Specimen-1 (from Assumed Hysteresis Loops)
180
:
....... -....... -
.... -
;,.,--
-----I
-
-Mander's model
;,......"""
-
-
- Kunnath's model
//
-
2/28
y
:
/
r
/
~ .. -- -
i-- ..
~
~ ' "
:
y'
.......
j
/ "
/
/;
I
;/ ;
y
J;
1/
1.8
1.6
1.4
1.2
x
<IJ
'tl
1
.5
<IJ
Cl
ro
0.8
E
ro
a
0.6
0.4
0.2
o
o 1 2 3 4 5 6 7 8 9 10 11 12 13
. No. of Loading Cycle
Figure-7.5.3 Damage Index of Specimen-2 (from Moment-Curvature Analysis)
-...--------------------r 180000
....... Cumulative
Dissipated
Energy
45000
40000
<D
C3
35000 >.
o
.s::
o
30000
Cll
UJ
l-
.E E
25000
>. E
e>z
20000
<D .:.::
c .....,
UJ
\j
15000 <D
....
Cll
0-
'00
10000
fI)
(5
5000
0
Dissipated
Energy
160000
140000 e>
<D
c
120000 UJ
\j
2
100000 [ E
'00 E
.!Q Z
80000 0.:.::
~ . . . . . ,
60000 ~
:5
E
40000 8
20000
o-, -,
>-- -, c-, c-, c-, c-, -, t>, c-, c-,
'D 'D 'D 'D 'D 'D 'D 'D 'D 'D 'D 'D
0J 0J -er- -e- co co co
= s s ~
0J
------
------
------
------
0J 0J 0J 0J
------
<;
Loading Cycle
0J 0J
Figure-7.5.4 Energy Dissipation of Specimen-2 (from Assumed Hysteresis Loops)
181
I
11
.
:
v
/
I
- - Mander's model
- - Kunnath's model
: : :: / : ,1
____ : __ . . ;Z . \ _
: , , , , , , : : : Y' , V' , ,
o : : : ,: : :
o 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
1.6
1.4
1.2
x 1
Ql
"0
.E
Ql
0.8
OJ
(\)
E
(\)
0.6 0
0.4
0.2
No. of Loading Cycle
Figure-7.5.5 Damage Index of Specimen-3 (from Moment-Curvature Analysis)
200000 >-
2l
CD
c:
UJ
"0
150000 2
ro-
.9- E
E
100000 '"'"
"5
E
::l
50000 ()
-r-------- -...,- 250000 50000
45000
<l.l
c::::::J Dissipated ts
40000
>-
Energy o
.c
35000
o
ro
-+- Cumulative
UJ
30000 L..
Dissipated
0-
- E
>- E
25000
Energy
e>z
<l.l
c: '"'"
20000 UJ
"C
<l.l
15000
.....
ro
a.
'00
10000
I/J
15
5000
0
Loading Cycle
Figure-7.5.6 Energy Dissipation of Specimen-3 (fromAssumed Hysteresis Loops)
182
I I I I I I I I I I
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
No. of Loading Cycle
2
1.8
1.6
1.4
x
(\)
1.2
"0
c
(\)
1
en
CIl
E
0.8 CIl
0
0.6
0.4
0.2
a
a
- - Mande's model
- - Kunnath's model
/
Figure-7.5.7 Damage Index of Specimen-4 (from Moment-Curvature Analysis)
60000 -,----------------...... 250000
50000
(\)
Dissipated
200000
u
>.
>.
Energy
en
o
...
(\)
.c
40000
c
L)
LU
CIl
.........Cumulative
150000 2
LU
...
Dissipated 0
E
CIl-
'+-
,9- E
>. E
30000
Energy
(f) E
~ z ,!Q Z
(\) ~
o ~
c'-"
100000 ~ '-"
LU
"0 :;:;
(\)
20000
CIl
.....
"5 CIl
0-
E
'iii
::::l
(/)
o
0
10000
50000
o
c-, c-, .>,
-0 -0 -0
-er- CD CD
----- -----
C'-J C'-J
Loading Cycle
Figure-7.5.8 Energy Dissipation of Specimen-4 (fromAssumed Hysteresis Loops)
183
/
/
/
9 10 11 12 13 8 7 6 5 4 3 2
.1 u
: / j
: - : ;:..
, ' : : : : , , y ,u ::,L--'--
" f _ ..-== .. r--- I I I I I 1 1
I I 1 I I I I I I
1
o --L.,
o
0.2
0.4
1.8
1.6
:
- -- -
:
:
1.4
--
-
- Mander's model
1.2
--
-
.
Kunnath's model
No. of Loading Cycle
Figure-7.5.9 Damage Index of Specimen-5 (from Moment-Curvature Analysis)
30000 -,------------------.... 160000
20000
>.
120000 e>
<I)
c
w
100000 -g
.....
co-
.9- E
80000 :g E
60000
:;::;
co
:;
40000
(J
140000
c:::J Dissipated
Energy
........... Cumulative
Dissipated
Energy
5000
25000
20000
15000
10000
<I)
(3
>.
(J
.c
o
co
W
Loading Cycle
Figure-7.5.10 Energy Dissipation of Specimen-5 (fromAssumed Hysteresis Loops)
...
0-
- E
>. E
e>z
<I)
c ......-
W
"0
<I)
ro
Q.
'iii
rn
i:S
184
The damage index of Specimen-1 calculated using the Kunnath's model predicts the
last load cycle of Specimen-1 as 2/12b)., beforethe ultimate state, as seen in Fig-7.5.1. And
the cumulative dissipated energy until 2/12b)., was calculated as 200,800kN'mm (Fig-7.5.2),
compared to 185,400kN'mm in the actual test (Fig-7.3.1).
The ultimate state cannot be applied to Specimen-2, however, the failure of the
Specimen-2 can be decided as to be during the first cycle to 10b)." when the initial cracks
occurred on the main-bars. The dissipated energy during this load cycle was
15,600kN'mm in the test (Fig-7.3.2), which is only 61% of that during the first cycle to 10b).,
of Specimen-1, 25,600kN'mm (Fig-7.3.1), showing that the specimen had already been
seriously damaged before that load cycle. From the analytical results, Specimen-2
approaches closely to failure but does not fail during the test (Figs-7.5.3 and 7.5.4)
according to Kunnath's model, but the Mander's model predicts the failure well. This
implies that the failure of Specimen-2, which was loaded with the extremely large first
displacement, was ruled by almost only by the fatigue of the main-bars with a very small
effect of confinement failure. Therefore, Mander's model might fit extreme loading cases.
The damage index of Specimen-3 exceeded 1.0 before the first load cycle to 1Ob)., was
completed and the cumulative dissipated energy exceeded 200,800kN'mm (which
corresponds to the ultimate state of Specimen-1) before the second load cycle to 8b)., was
completed. Therefore, it was decided that the last load cycle before the ultimate state was
taken as the first cycle to 8b)." which is equal to the experimental result (Figs-7.5.5 and
7.5.6).
For Specimen-4, the last load cycle before the ultimate state was chosen as the
second cycle to 8b)." which is also same as the experimental result (Figs-7.5.7 and 7.5.8).
The analytical result of Specimen-5 is recognised asthe standard case for the columns
with the weak concrete, and not related to the other results (Figs-7.5.9 and 7.5.10). The
predicted energy dissipation capacity 145,200kN'mm (2/10b).,) is 9% larger than
132,800kN'mm (1/10b).,) obtained in the test. The proposed procedure predicted well the
failure of Specimen-5 even though its concrete strength was extremely low. This result
indicates that the proposed failure-prediction procedure covers a range of concrete
strength, from extremely low to more normal.
185
Figs-7.5.11 to 7.5.15 inclusive show the assumed hysteresis loops for the test
specimens. The modified Takeda model (Otani [29]) was used with a=0, p:=0 and FO (seen
in Fig-7.5.16). The lateral strength of Specimen-1 for the second load cycle to the same
displacement was assumed to be 90% of that for the first cycle. For Specimen-2, no
strength degradation was assumed for the loops enclosed in any other loops. In the
analysis of Specimen-3, the additional displacement explained in 6.3.3.3 was considered
in the calculated displacement using the moment-curvature analysis. The lateral strength
for the second load cycle in the east-west direction was assumed to be 90% of that for the
first load cycle, and 80% and 70%, for the first load cycle and the second load cycle in the
north-south direction, respectively. For Specimen-4, the first cycle to the same
displacement was the first cycle in the east-west direction, the second was the first cycle in
the north-south direction, the third was the second cycle in the east-west direction and the
fourth was the second cycle in the north-south direction. The lateral strength of the column
was decreased by 10% in each load cycles to the same displacement.
186
Displacement Ductility Factor
-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12
200 -r-----t---t--+--+--+--l--t----t--l--t---+--+--I-------,
150 100 50 o -50 -100
50
-50
100
150
-100
-150
-200 +-------+-----+-----+-----+---'---+------1
-150
Z
::::-
"'0
ro
o
...J
~
(1)
10
...J
Displacement (mm)
Figure-7.5.11 Assumed Hysteresis Loops for Specimen-1
Displacement Ductility Factor
-12 -10 -8 -6 -4 -2 0 2 4 6 8 10 12
....-: , -/ , .
-- _. - .. ----- - - - - - --- - - ~ - -- - - - - ~ --- ---- _. - - - - - ---
/. . _ ~ ::.;:.-i .
, ,
-150 ~ : -- 1/ ~ -; .
200
150
100
Z 50
::::-
"'0
ro
0
0
...J
(ij
....
(1)
10 -50
...J
-100
++-------+------+
150 100 50 o -50 -100
-200 -1-----+-----+----+-----+----+-----1
-150
Displacement (mm)
Figure-7.5.12 Assumed Hysteresis Loops for Specimen-2
187
-10 -8
Displacement Ductility Factor
4 0 2 4 6 8 10
200 ....----.,..----r--;---r--.,---..--;---.,..--.,.----,---.,..---.
150
100
Z 50
C
"0
ro
.3
Q)
ro -50
...J
-100
-150
150 100 50 o -50 -100
-200 .L- --+ -+- -I- -+ -+- ---!
-150
Displacement (mm)
a) East-West Direction
-8 -6
Displacement Ductility Factor
4 -2 0 2 4 6 8
150 -,.---..,.----r------,----.,...---.,...----r------r---.,...-----r--.
100
"0
ro
o
...J
Q)
ro
...J
50
-50
-100
100 80 60 40 20 o -20 40 -60 -80
-150 ---'-----;---+---+------!'-----+---+---+---+---""'"'
-100
Displacement (mm)
b) North-South Direction
Fig-7.5.13 Assumed Hysteresis Loops for Specimen-3
188
Displacement Ductility Factor
-10 -8 -6 -4 -2 0 2 4 6 8 10
200
150
100
Z
50
6
"'0
1Il
0
0
-l
ro
...
Q)
-1Il
-50
-l
-100
-150
-200
-100 -80 -60 -40 -20 0 20 40 60 80 100
Displacement (mm)
a) East-West Direction
Displacement Ductility Factor
-10 -8 -6 -4 -2 0 2 4 6 8 10
200
150
100
Z 50
:::..
"'0
ro
0
0
-l
"
Q)
co
-50
-l
-100
-150
-200
-100 -80 -60 -40 -20 0 20 40 60 80 100
Displacement (mm)
b) North-South Direction
Fig-7.5.14 Assumed Hysteresis Loops for Specimen-4
189
Displacement Ductility Factor
-10 -8 -6 -4 -2 0 2 4 6 8 10
200 -....- --=-'1----+---+_--1-_-+-'-_-;-_-1-_-1-_-1--_-1--_+-- ---,
150
100
Z 50
-100
-150
150 100 50 o
Displacement (mm)
-50 -100
-200 -!-----i------+-----+-----+-----+-------\
-150
Figure-7.5.15 Assumed Hysteresis Loops for Specimen-5
F
B
Ku :: K
u
(1<0 .c )
a= o,Ku=Ko
Increasing a softens unloading
K
L
=KL'(K
o
I
P:: oKL points to 8
o-IKL tends /0 point to A
Increa.sing 0.stiffens reloading
Figure-7.5.16 Modified Takeda Model (Otani [29])
...... 6
Note:
-
Clough
a-Hyst
190
Chapter VUI
Discussion and Conclusion
8.1 General
The findings from the experimental data and the analysis carried out concerning plastic
hinge zone length, lateral strength-curvature relationship, energy dissipation and damage
accumulation, are listed and discussed, and some conclusions are reached. These
findings and conclusions are specifically applicable only for the square or rectangular
reinforced concrete bridge columns which are designed using the latest Japanese
specifications with low axial load and which fail by flexure.
8.2 Research Findings
8.2.1 Findings concerning Plastic Hinge Zone Length
1) Plastic hinge zone length L
p
tends to be small for the displacement ductility factor us at
around 2; increases as Il.d increases, and at 1l.d=4, becomes stable at about the
theoretical values calculated using the formulas suggested by Paulayand Priestley [17]
and the Japanese Specification.
2) The analysed experimental data and literature review confirmed that the contribution of
the yield penetration of the vertical main-bars on the plastic hinge zone length is small at
first and increases suddenly when the strain reaches the onset of the strain hardening,
to a certain level which is in proportion to the displacement of the column.
3) The stabilised plastic hinge zone length L
p
is not affected by the cyclic loading pattern,
but could be affected by the concrete strength.
8.2.2 Findings concerning Moment-Curvature Analysis
1) The moment-curvature analysis suggested by Mander et al. [23] predicts well the
maximum lateral strength of the test specimens.
2) The lateral strength-displacement relationship obtained using the moment-curvature
191
analysis has a plateau from the yielding point to the onset of strain hardening of the
main-bars. On the other hand, the experimental data shows no plateau and that the
strain in main-bar in tension suddenly increased just after it yielded. This difference can
be minimised by assuming a pseudo stress-strain relationship for the main-bars in the
analysis, in which the onset of the strain hardening is at the yield point.
3) The lateral strength of the column starts decreasing once the main-bars buckle in the
load cycle leading to the confinement failure. The lateral strength decreases also by
repeating the load cycle to the same displacement even before buckling of the main-
bars, if the maximum displacement of the load cycle has not been exceeded before.
These effects of cyclic loading are not reflected in the moment-curvature analysis, and
that makes it difficult to predict the behaviour of the specimen in the later stage of cyclic
loading test.
8.2.3 Findings concerning Energy Dissipation
1)AII the test specimens except Specimen-2 sustained good energy dissipation until the
displacement ductility factor JlA reached 6 which is equal to the design available ductility
factor of the prototype column designed using the Japanese Specification.
2)The dissipated energy by Specimens-1, 3 and 4 are approximately the same. This
suggests that the energy dissipation capacity of the test specimens does not matter if the
cyclic loading is uni-directional or bi-directional, if the displacement amplitude is
increased step by step as used for Specimens-1, 3 and 4.
3)The energy dissipation capacity of a reinforced concrete column depends on when the
buckling of the main-bar and the confinement failure occur.
4)The loading pattern and the concrete strength can affect energy dissipation as seen in
the experimental results of Specimens-2 and 5, respectively.
8.2.4 Findings concerning Damage Index
1)The calculated damage index using a fatigue-based damage model suggested by
Kunnath et al. [6] (based on an experimental study with some reinforced concrete
columns which were seismically detailed using CALTRANS or MSHTO) predicts well the
failure of the f1exurally reinforced concrete column. These columns carried low axial
loads, had common concrete strengths, and were seismically detailed using the
Japanese specification. As well, the standard loading pattern (Park [2]) in one direction
was used in the test. It also predicts well the failure of a column loaded independently in
192
two directions perpendicular to each other, calculating the damage index for the corner-
bars which take the most serious damage.
2)Mander's model predicted well the initial cracking of the main-bars of Specimen-2.
3)Loading pattem and concrete strength affect the prediction of the failure of a reinforced
concrete column by the fatigue based damage models.
8.2.5 Findings concerning Proposed Failure Prediction Procedure
The analytical result using the proposed procedure for failure prediction corresponded
well to the experimental result of Specimens-1, 3, 4 and 5. However, both the Kunnath's
model and the dissipated energy seemed to underestimate the failure of Specimen-2,
which was subjected an unusual loading pattern which was opposite to that used for
Specimen-i.
8.3 Discussion
8.3.1 Discussion concerning Plastic Hinge Zone Length
As described in 8.2.1, the plastic hinge zone length L
p
obtained from the experimental
data could be shorter than the theoretical values suggested by Paulay and Priestley [17]
and the Japanese specification when the displacement ductilities are less than 4. In
5.3.3.2(b) for instance, the plastic hinge zone length L
p
of the prototype column in the
transverse direction was 2.0m, and from it, the allowable ductility factor Jla was obtained as
1.74 for the Type-1 earthquake. Using the proposed Lp-Jl.1relationship seen in Fig-7.1.14,
L
p
and Jla become 0.43m and 1.14, respectively. These values were obtained by repeating
the calculation equilibrating L
p
and Jl1 corresponding to au, becalJsethe ultimate
displacement au is obtained using L
p
with Eq(4.4.13), Jl1 is calculated frorri au, and then L
p
is
obtained using Fig-7.1.14. As a result, the design lateral selsrniccoefflclentjc; increases
from 0.54 to 0.75 and means that the proposed L
p
-JlL1 relationship increases the design
lateral seismic coefficient k
hc
of the prototype column by 40%. Overestimation of plastic
hinge zone length should be avoided. As mentioned in 8.2.1(1), the more cyclic loading
applied to the column, the more L
p
increases and is stabilised at around the theoretical
value of Paulay and Priestley [17]. However, the possible smallest value for L
p
should be
used in design because the number of critical seismic waves in the event of a large
earthquake is usually limited and might come at the beginning of the earthquake as seen in
193
the seismic records of Hyogo-ken Nanbu earthquake.
8.3.2 Discussion concerning Moment-CurvatureAnalysis
The moment-curvature analysis can predict well the lateral strength-displacement
relationship of a reinforced concrete bridge column which is monotonically loaded,
especially when the pseudo stress-strain relationship of the main-bars in which the onset of
strain hardening is adjusted, suggested in 7.2, is used. However, the envelope of
hysteresis loops declines much earlier than the lateral strength-displacement relationship
obtained using the moment-curvature analysis, due to the main-bar-failure and confinement
failure. This is because the moment-curvature analysis assumes that the main-bars are
well supported and do not buckle, but on the other hand, the main-bars in a column
designed using the Japanese specification buckle early due to much less number of cross-
bars than that required by the New Zealand specification, CALTRANS, AASHTO or Euro
Code. This behaviour can still be predicted if main-bar-buckling is considered as well as
the adjustment of onset of strain hardening in the stress-strain relationship of the main-
bars, and the confinement failure is considered in the stress-strain relationship of the core-
concrete. However, more experimental studies are necessary to include the main-bar-
buckling and confinement failure in the moment-curvature analysis. Especially, the
prediction of main-bar-buckling in a column designed using the Japanese specification is a
complex issue, because the restriction on a main-bar by transverse reinforcement against
buckling is assumed to depend on the distance between the bar and the nearest corner of
a hoop or cross-bar which holds the hoop with its hook.
8.3.3 Discussion concerning Energy Dissipation
As far as Specimens-1, 3 and 4 are concerned, energy dissipation is a good option to
predict a failure of the test specimens, because it is independent of whether the cyclic
loading is in uni-direction or bl-directlon, provided that the displacement amplitude of a load
cycle starts from a small value and increases step-by-step like the standard loading pattern
suggested by Park [2]. Specimens-2 and 5, however, show that the energy dissipation
capacity can be affected by the loading pattern and concrete strength. The maximum
displacement ductility factors of Specimens-3 and 4 which were reached before the ultimate
state in the tests, were less than that of Specimen-1, but greater than the design available
ductility factor. It means that the maximum displacement ductilities, which a reinforced
concrete column subjected to a bi-directional cyclic loading can reach, is less than that of
194
the same column with the standard uni-directional cyclic loading, but the safety factor a in
Eq(4.4.12), defined in Table-4.4.3, can absorb the effect of the bi-directional loading
patterns.
8.3.4 Discussion concerning Damage Index
Kunnath's damage model predicted the failure of Specimens-1 and 3 well, but
underestimated the failure of Specimen-4, although the dissipated energy by Specimen-4
was approximately the same as that of Specimens-1 and 3. Mander's model, which was
derived from the damage only on main-bars and always reaches to 0=1 earlier than that by
Kunnath's model, but also underestimated the failure of Specimen-4, as for Kunnath's
model. These findings imply that the loading pattern applied to Specimen-4 caused less
damage to the main-bars but achieved the same energy dissipation on the whole column
as that of Specimens-1 and 3, until the ultimate state was reached. Mander's model
predicted well the initial crack of the main-bars of Specimen-2 and it seems to correspond
the actual failure of Specimen-2 based on the energy dissipation, instead of the ultimate
state defined by Zahn et al. [18] which could not applied due to the unusual loading pattern.
The failure of Specimen-5 was underestimated by Kunnath's model even though it
predicted the failure of Specimen-1 well which was subjected to the same loading pattern.
These findings and assumptions above show that the failure of a reinforced concrete
column results from both the main-bar failure and confinement failure, and the balance of
them is affected by the loading patternapplied to the specimen and the concrete strength.
8.3.5 Discussion concerning Proposed Failure Prediction Procedure
It was confirmed that the proposed procedure predicts well the failure of reinforced
concrete columns seismically detailed by the Japanese specification when the column is
subjected to a quasi-static cyclic loading with the displacement amplitude is increased
gradually. On the other hand, when the column is SUbjected to an extremely large
displacement at the beginning of the cyclic loading, the damage to the main-bars is much
greater than that of the transverse reinforcement and Mander's model agrees well with the
failure of the column. Therefore, the applied damage model should be decided depending
on the loading pattern.
195
8.4 Conclusion
1. Shorter plastic hinge zone lengths should be used in the seismic design of f1exurally
reinforced rectangular concrete bridge columns with low axial stress using the Japanese
specification, when the ductility factor corresponding the ultimate displacement of the
column is less than four. A plastic hinge zone length-displacement ductility factor
relationship, seen in Fig-7.1.14, was proposed for the Japanese specification.
2. If an extremely large displacement, such as ,u.d=12, for the specimens used in this
research, is applied to a column at the early stage of a cyclic loading, it may lead to the
buckling of the main-bars and confinement failure with only small energy dissipation.
However, as long as the displacement amplitude in the cyclic loading starts at a small
level and increases step-by-step, like the standard loading pattern suggested by Park [2],
the energy dissipation capacity of a column until the ultimate state is the same for both
uni-directional and bi-directionalloading.
3. The maximum displacement of a column when it reaches the ultimate state in a bi-
directional cyclic loading, is smaller than that of the same column subjected to the
standard uni-directional loading pattern suggested by Park [2]. However, the design
available displacement ductility factor was met because of a safety factor defined in the
Japanese specification.
4. The failure of a column results from either the main-bar failure or confinement failure,
and the balance of them is affected by the loading pattern applied to the specimen and
the concrete strength.
5. The fatigue based damage model suggested by Kunnath et al. [6] predicted well the
ultimate state of a column when the fatigue of the critical main-bars is relatively greater
when compared with the dissipated energy.
6. The failure of a column corresponds well to the cumulative dissipated energy calculated
for the same column cyclically loaded by the standard loading pattern suggested by Park
[2], when the fatigue of the critical main-bars is relatively low.
7. Based on the findings in 5 and 6, a procedure to predict the failure of reinforced concrete
196
columns in arbitrary cyclic loading patterns was suggested, as described in 7.5. The
result of failure prediction carried out using the suggested procedure corresponded well
to the experimental result, except only Specimen-2, which was loaded with extremely
large first displacement. For some extreme loading patterns, other proper damage
models should be chosen.
8.5 Recommendations for Future Research
The conclusion stated in this report is applicable only for limited situation. The
following recommendations for future work should provide comprehensive information for
the damage estimation of reinforced concrete bridge columns.
1) Bridge columns with a much smaller aspect ratio should be tested using basically the
same procedure of this research. This should provide useful information concerning the
fatigue of a column with a shear failure mode, and can confirm the effect of shorter plastic
hinge length with small allowable ductility factor. The axial load also should be varied.
2) A circular column, which provides more efficient confinement to the core concrete, is
recommended to be tested.
3) A rectangular reinforced concrete column with the same dimension and seismically
detailed by some other specifications (AASHTO, CALTRANS or New Zealand code)
should be tested and compared to the result of this research.
4) Some random cyclic loading patterns bi-directionally simulating actual earthquakes are
recommended to be used by quasi-static loading or shaking table.
5) The effect of an eccentric axial load on the fatigue of columns should be investigated.
The rotation caused by a large eccentricity of axial load or a curved superstructure might
have some effect on the fatigue of column.
6) A simple fatigue test of a reinforcing bar with a longer effective buckling length should be
carried out for the main-bars either at a corner of a hoop or adjacent to a cross-bar in a
column designed using the Japanese specification.
197
List of References
[1] Japan Road Association, "Guide Specification of Seismic Design of Highway Bridges"
(in Japanese), 1996.
[2] Park, R. "Evaluation of Ductility of Structures and Structural Assemblages from
laboratory Testing", Bulletin of the New Zealand National Society for Earthquake
Engineering, Vol.22, No.3, September, 1989.
[3] .Minutes of the Second US-New Zealand-Japan-China seminar on Earthquake
Resistant Design of Ductile Reinforced Concrete Frame with Emphasis on Reinforced
Concrete Beam-Column Connections, Tokyo, 1985.
[4] Cheok, G.S. and Stone, W.C., "Behavior of 1/6-Scale Model Bridge Columns Subjected
to Cyclic Inelastic Loading", National Bureau of Standards, U.S. Department of
Commerce, November, 1986
[5] Stone, W.C. and Cheok, G.S., "Inelastic Behavior of Full-Scale Bridge Columns
Subjected to Cyclic Loading", National Bureau of Standards and Technology, U.S.
Department of Commerce, January, 1989
[6] Kunnath, S.K., EI-Bahy, A., Taylor, A.W. and Stone, W.C., "Cumulative Seismic Damage
of Reinforced Concrete Bridge Piers", National Bureau of Standards and Technology,
Technology Administration, U.S. Department of Commerce, October, 1997.
[7] Kawashima, K., "Seismic Design Force of Bridges", Seismic Design and Retrofitting of
Reinforced Concrete Bridges, Proceedings of the Second International Workshop held
in Queenstown, New Zealand, August, 1994.
[8] Mori, T., Zhang, J. and Kawashima, K., "Experimental Study on the Effectiveness of
Lateral PC Tendons on Confinement of Reinforced Concrete Piers"(in Japanese),
Report No.TIT/EERG 98-8, Department of Civil Engineering, Tokyo Institute of
Technology, June, 1998.
198
[9] Klazig, W.B. and Meskouris, M., "Nonlinear Seismic Analysis of Reinforced Concrete
Frames", Earthquake Prognostics (Vogel and Brandes, editors), Verlag Friedr, Vieweg
and Sohn, Braunschweig, 1987.
[10] Park, Y.J. and Ang, A.H-S., "Mechanistic Seismic Damage Model for Reinforced
Concrete", Jounal of Structural Engineering, ASCE, 111, pp.722-739, 1985.
[11] Powell, G. and Allahabadi, R., "Seismic Damage Prediction by Deterministic Methods:
Concepts and Procedures", Earthquake Engineering and Structural Dynamics, Vo1.16,
1988.
[12] Williams, M.S. and Sexsmith, R.G., "Review of Methods Assessing Seismic Damage
in Concrete Structures", Technical Report 94-02, Earthquake Engineering Research
Facility, University of British Columbia, Canada, 1994.
[13] Mander, J.B. and Cheng, C.-T., "Renewable Hinge Detailing for Bridge Columns",
Pacific Conference on Earthquake Engineering, Melbourne, Australia, November 20-
22,1995.
[14] Coffin, L.F.Jr., "A Study of the Effects of Cyclic Thermal Stresses on a Ductile Metal",
Transactions of the American Society of Mechanical Engineers, New York, N.Y., 76,
931-950,1954.
[15] Manson, S.S., "Behavior of Materials under Conditions of Thermal Stress", Heat
Transfer Symposium, University of Michigan Engineering Research Institute, Ann
Arbor, Michigan, pp.9-75, 1953.
[16] Mander, J.B., Panthaki, F.D. and Kasalanati, A., "Low Cycle Fatigue Behavior of
Reinforcing Steel", Journal of Materials in Civil Engineering, ASCE, Vol.6, No.4, 1994.
[17] Paulay, T. and Priestley, M.J.N., "Seismic Design of Reinforced Concrete and
Masonry Buildings", John Wiley and Sons, New York, 1992.
[18] Zahn, F.A., Park, R. and Priestley, M.J.N., "Design of Reinforced Concrete Bridge
Columns for Strength and Ductility", Department of Civil Engineering, University of
Canterbury, New Zealand, March, 1986.
199
[19] Japan Road Association, "Seismic Design Data for Highway Bridges"(in Japanese),
1997.
[20] Wong, Y.L., Paulay, T., Park, R and Priestley, M.J.N, "Squat Circular Bridge Piers
under Multi-Directional Seismic Seismic Attack", Department of Civil Engineering,
University of Canterbury, New Zealand, March, 1990.
[21] NZS 3101, "Concrete Structures Standard, Part-1:The Design of Concrete Structures",
Standards New Zealand, Wellington, 1995
[22] Sato, Y., Tanaka, H. and Park, R, "Reinforced Concrete Columns with Mixed Grade
Longitudinal Reinforcement", Department of Civil Engineering, University of
Canterbury, August, 1993.
[23] Mander, J.B., Priestley, M.J.N. and Park, R, "Seismic Design of Bridge Piers",
Department of Civil Engineering, University of Canterbury, August, 1984.
[24] Shima, H., Syu, L. and Okamura, H., "Bond-Slip-Strain Relationship of Deformed Bars
Embedded in Massive Concrete"(in Japanese), Proceedings of JSCE/No.378, Japan
Society of Civil Engineers, Tokyo, February, 1987.
[25] Ogasawara, M. and Tsuno, K., "An Experimental Study Concerning Tie-bars of
Reinforced Concrete Piers", Proceedings of 9
th
REAAA Conference held in Wellington,
New Zealand, 1998
[26] Metropolitan Expressway Public Corporation, "Research Report concerning Highway
Bridges"(in Japanese), Concrete Structure Group, 1996.
[27] Ohno, T. and Nshioka, T., "An Experimental Study on Energy Absorption Capacity of
Columns in Reinforced Concrete Structures", Structural Engineering/Earthquake
Engineering, Proceedings of the Japan Society of Civil Engineers, Vol.1, No.2,
October, 1984.
[28] Kunnath, S.K. and Reinhorn, A.M., u Efficient Modelling Scheme for Transient Analysis
of Inelastic RC Structures", Microcomputer in Civil Engineering, Vol.10, No.1, pp.97-
200
110, 1995.
[29] Otani, S., "A Computer Program for Inelastic Response of RIC Frames to
Earthquakes", Report UILU-Eng-74-2029, Civil Engineering Studies, University of
Illinois at Urbana-Champaign, November, 1974.
201